vecId
stringlengths
12
23
id
stringlengths
2
13
conference
stringclasses
11 values
year
float64
2.02k
2.03k
title
stringlengths
6
189
abstract
stringlengths
10
4.74k
author
stringlengths
0
7.45k
aff
stringlengths
0
7.16k
status
stringclasses
11 values
track
stringclasses
4 values
keywords
stringlengths
0
804
github
stringlengths
0
141
site
stringlengths
0
193
gsCitation
float64
-1
11.1k
arxiv
stringlengths
0
12
text
stringlengths
58
4.82k
vector
list
wacv_2025_b0f716002e
b0f716002e
wacv
2,025
Metric Compatible Training for Online Backfilling in Large-Scale Retrieval
Backfilling is the process of re-extracting all gallery embeddings from upgraded models in image retrieval systems. It inevitably spends a prohibitively large amount of computational cost and even entails the downtime of the service. Although backward-compatible learning sidesteps this challenge by tackling query-side ...
Seonguk Seo; Mustafa Gokhan Uzunbas; Bohyung Han; Sara Cao; Ser-Nam Lim
ECE & IPAI, Seoul National University; Meta; ECE & IPAI, Seoul National University + Meta; Meta; University of Central Florida
Poster
main
https://openaccess.thecvf.com/content/WACV2025/html/Seo_Metric_Compatible_Training_for_Online_Backfilling_in_Large-Scale_Retrieval_WACV_2025_paper.html
3
2301.03767
Metric Compatible Training for Online Backfilling in Large-Scale Retrieval Backfilling is the process of re-extracting all gallery embeddings from upgraded models in image retrieval systems. It inevitably spends a prohibitively large amount of computational cost and even entails the downtime of the service. Although ba...
[ -0.05939887464046478, -0.037915781140327454, -0.03806653991341591, -0.00024174372083507478, -0.004671160131692886, -0.04157167673110962, 0.03365685045719147, 0.017186477780342102, -0.01005371194332838, 0.01758221723139286, -0.047790467739105225, 0.004086970817297697, 0.005889006424695253, ...
wacv_2025_c3c2ea451f
c3c2ea451f
wacv
2,025
MimicGait: A Model Agnostic Approach for Occluded Gait Recognition using Correlational Knowledge Distillation
Gait recognition is an important biometric technique over large distances. State-of-the-art gait recognition systems perform very well in controlled environments at close range. Recently there has been an increased interest in gait recognition in the wild prompted by the collection of outdoor more challenging datasets ...
Ayush Gupta; Rama Chellappa
Johns Hopkins University; Johns Hopkins University
Poster
main
https://github.com/Ayush-00/mimicgait
https://openaccess.thecvf.com/content/WACV2025/html/Gupta_MimicGait_A_Model_Agnostic_Approach_for_Occluded_Gait_Recognition_using_WACV_2025_paper.html
0
2501.15666
MimicGait: A Model Agnostic Approach for Occluded Gait Recognition using Correlational Knowledge Distillation Gait recognition is an important biometric technique over large distances. State-of-the-art gait recognition systems perform very well in controlled environments at close range. Recently there has been an incre...
[ -0.0676504448056221, -0.037023380398750305, -0.04176418110728264, 0.02799328975379467, -0.028651732951402664, -0.057980723679065704, -0.022029664367437363, 0.001997672952711582, 0.021201906725764275, 0.03010031022131443, -0.04225331172347069, -0.05064377188682556, -0.008804340846836567, 0....
wacv_2025_40f2127f95
40f2127f95
wacv
2,025
Mind the Map! Accounting for Existing Maps When Estimating Online HDMaps from Sensors
While HDMaps are a crucial component of autonomous driving they are expensive to acquire and maintain. Estimating these maps from sensors therefore promises to significantly lighten costs. These estimations however overlook existing HDMaps with current methods at most geolocalizing low quality maps or considering a gen...
Rémy Sun; Li Yang; Diane Lingrand; Frederic Precioso
Universit ´e Cˆote d’Azur, Inria, CNRS, I3S, Maasai, Nice, France; Universit ´e Cˆote d’Azur, Inria, CNRS, I3S, Maasai, Nice, France; Universit ´e Cˆote d’Azur, Inria, CNRS, I3S, Maasai, Nice, France; Universit ´e Cˆote d’Azur, Inria, CNRS, I3S, Maasai, Nice, France
Poster
main
https://openaccess.thecvf.com/content/WACV2025/html/Sun_Mind_the_Map_Accounting_for_Existing_Maps_When_Estimating_Online_WACV_2025_paper.html
0
Mind the Map! Accounting for Existing Maps When Estimating Online HDMaps from Sensors While HDMaps are a crucial component of autonomous driving they are expensive to acquire and maintain. Estimating these maps from sensors therefore promises to significantly lighten costs. These estimations however overlook existing H...
[ -0.05788809806108475, -0.0055661629885435104, -0.024098211899399757, 0.001899043912999332, -0.0013272258220240474, -0.03549247607588768, 0.05594227835536003, 0.008611181750893593, 0.029748570173978806, 0.02106722630560398, -0.006908590905368328, -0.014602992683649063, 0.008802956901490688, ...
wacv_2025_cc9bb420f7
cc9bb420f7
wacv
2,025
Mind the Prompt: A Novel Benchmark for Prompt-Based Class-Agnostic Counting
Recently object counting has shifted towards class-agnostic counting (CAC) which counts instances of arbitrary object classes never seen during model training. With advancements in robust vision-and-language foundation models there is a growing interest in prompt-based CAC where object categories are specified using na...
Luca Ciampi; Nicola Messina; Matteo Pierucci; Giuseppe Amato; Marco Avvenuti; Fabrizio Falchi
CNR-ISTI, Pisa, Italy; CNR-ISTI, Pisa, Italy; University of Pisa, Italy; CNR-ISTI, Pisa, Italy; University of Pisa, Italy; CNR-ISTI, Pisa, Italy
Poster
main
https://github.com/ciampluca/PrACo
https://openaccess.thecvf.com/content/WACV2025/html/Ciampi_Mind_the_Prompt_A_Novel_Benchmark_for_Prompt-Based_Class-Agnostic_Counting_WACV_2025_paper.html
2
2409.15953
Mind the Prompt: A Novel Benchmark for Prompt-Based Class-Agnostic Counting Recently object counting has shifted towards class-agnostic counting (CAC) which counts instances of arbitrary object classes never seen during model training. With advancements in robust vision-and-language foundation models there is a growing...
[ -0.04906918853521347, -0.026971563696861267, -0.026201995089650154, -0.021144825965166092, 0.01300022006034851, -0.05438288301229477, 0.005620602518320084, -0.0007369311060756445, 0.0004683835431933403, 0.02922530099749565, -0.015400542877614498, -0.04489152878522873, -0.02242744155228138, ...
wacv_2025_dfd048130d
dfd048130d
wacv
2,025
MissionGNN: Hierarchical Multimodal GNN-Based Weakly Supervised Video Anomaly Recognition with Mission-Specific Knowledge Graph Generation
In the context of escalating safety concerns across various domains the tasks of Video Anomaly Detection (VAD) and Video Anomaly Recognition (VAR) have emerged as critically important for applications in intelligent surveillance evidence investigation violence alerting etc. These tasks aimed at identifying and classify...
Sanggeon Yun; Ryozo Masukawa; Minhyoung Na; Mohsen Imani
University of California, Irvine; University of California, Irvine; Kookmin University; University of California, Irvine
Poster
main
https://openaccess.thecvf.com/content/WACV2025/html/Yun_MissionGNN_Hierarchical_Multimodal_GNN-Based_Weakly_Supervised_Video_Anomaly_Recognition_with_WACV_2025_paper.html
7
2406.18815
MissionGNN: Hierarchical Multimodal GNN-Based Weakly Supervised Video Anomaly Recognition with Mission-Specific Knowledge Graph Generation In the context of escalating safety concerns across various domains the tasks of Video Anomaly Detection (VAD) and Video Anomaly Recognition (VAR) have emerged as critically importa...
[ -0.06721660494804382, -0.04929697513580322, 0.0094995629042387, -0.00511860940605402, -0.00928366370499134, -0.005874256603419781, -0.004866727162152529, 0.00969747081398964, -0.011316714808344841, 0.00939161330461502, -0.02826479822397232, -0.024810412898659706, -0.013871520757675171, 0.0...
wacv_2025_e3e83c23fb
e3e83c23fb
wacv
2,025
MixDiff: Mixing Natural and Synthetic Images for Robust Self-Supervised Representations
This paper introduces MixDiff a new self-supervised learning (SSL) pre-training framework that combines real and synthetic images. Unlike traditional SSL methods that predominantly use real images MixDiff uses a variant of Stable Diffusion to replace an augmented instance of a real image facilitating the learning of cr...
Reza Akbarian Bafghi; Nidhin Harilal; Maziar Raissi; Claire Monteleoni
University of Colorado, Boulder; University of Colorado, Boulder; University of Colorado, Boulder+INRIA, Paris; University of California, Riverside
Poster
main
https://github.com/cryptonymous9/mixing-ssl
https://openaccess.thecvf.com/content/WACV2025/html/Bafghi_MixDiff_Mixing_Natural_and_Synthetic_Images_for_Robust_Self-Supervised_Representations_WACV_2025_paper.html
0
2406.12368
MixDiff: Mixing Natural and Synthetic Images for Robust Self-Supervised Representations This paper introduces MixDiff a new self-supervised learning (SSL) pre-training framework that combines real and synthetic images. Unlike traditional SSL methods that predominantly use real images MixDiff uses a variant of Stable Di...
[ -0.06700655072927475, -0.03401477634906769, -0.023821303620934486, -0.01750975474715233, -0.003959560766816139, 0.006644937675446272, 0.03986049443483353, 0.02869882434606552, 0.025501947849988937, 0.026305733248591423, -0.007686206605285406, -0.04205263778567314, -0.05659386143088341, 0.0...
wacv_2025_cf0af5064b
cf0af5064b
wacv
2,025
Mixed Patch Visible-Infrared Modality Agnostic Object Detection
In real-world scenarios using multiple modalities like visible (RGB) and infrared (IR) can greatly improve the performance of a predictive task such as object detection (OD). Multimodal learning is a common way to leverage these modalities where multiple modality-specific encoders and a fusion module are used to improv...
Heitor R. Medeiros; David Latortue; Eric Granger; Marco Pedersoli
Laboratoire d’imagerie, de vision et d’intelligence artificielle (LIVIA) + International Laboratory on Learning Systems (ILLS) + Dept. of Systems Engineering, ETS Montreal, Canada; Laboratoire d’imagerie, de vision et d’intelligence artificielle (LIVIA) + International Laboratory on Learning Systems (ILLS) + Dept. of Sys...
Poster
main
https://github.com/heitorrapela/MiPa
https://openaccess.thecvf.com/content/WACV2025/html/Medeiros_Mixed_Patch_Visible-Infrared_Modality_Agnostic_Object_Detection_WACV_2025_paper.html
0
Mixed Patch Visible-Infrared Modality Agnostic Object Detection In real-world scenarios using multiple modalities like visible (RGB) and infrared (IR) can greatly improve the performance of a predictive task such as object detection (OD). Multimodal learning is a common way to leverage these modalities where multiple m...
[ -0.03551537171006203, -0.02848244272172451, 0.015699492767453194, 0.030642161145806313, -0.02479061670601368, -0.02628580667078495, 0.003677982371300459, 0.028925461694598198, 0.04766148328781128, 0.022797029465436935, -0.07261823117733002, -0.06593602150678635, 0.024052251130342484, -0.00...
wacv_2025_12da6b4778
12da6b4778
wacv
2,025
MoRAG - Multi-Fusion Retrieval Augmented Generation for Human Motion
We introduce MoRAG a novel multi-part fusion based retrieval-augmented generation strategy for text-based human motion generation. The method enhances motion diffusion models by leveraging additional knowledge obtained through an improved motion retrieval process. By effectively prompting large language models (LLMs) w...
Sai Shashank Kalakonda; Shubh Maheshwari; Ravi Kiran Sarvadevabhatla
CVIT, IIIT Hyderabad; University of California San Diego; CVIT, IIIT Hyderabad
Poster
main
https://openaccess.thecvf.com/content/WACV2025/html/Kalakonda_MoRAG_-_Multi-Fusion_Retrieval_Augmented_Generation_for_Human_Motion_WACV_2025_paper.html
1
MoRAG - Multi-Fusion Retrieval Augmented Generation for Human Motion We introduce MoRAG a novel multi-part fusion based retrieval-augmented generation strategy for text-based human motion generation. The method enhances motion diffusion models by leveraging additional knowledge obtained through an improved motion retri...
[ 0.007798853330314159, -0.02953190729022026, -0.010149256326258183, -0.002668151631951332, -0.022560134530067444, 0.008004454895853996, 0.018429405987262726, 0.022036783397197723, -0.020354585722088814, 0.006569914519786835, -0.03775598108768463, -0.030522534623742104, -0.03336357697844505, ...
wacv_2025_37226ba84e
37226ba84e
wacv
2,025
Modality-Incremental Learning with Disjoint Relevance Mapping Networks for Image-Based Semantic Segmentation
In autonomous driving environment perception has significantly advanced with the utilization of deep learning techniques for diverse sensors such as cameras depth sensors or infrared sensors. The diversity in the sensor stack increases the safety and contributes to robustness against adverse weather and lighting condit...
Niharika Hegde; Shishir Muralidhara; René Schuster; Didier Stricker
RPTU – University of Kaiserslautern-Landau + DFKI – German Research Center for Artificial Intelligence; RPTU – University of Kaiserslautern-Landau + DFKI – German Research Center for Artificial Intelligence; RPTU – University of Kaiserslautern-Landau + DFKI – German Research Center for Artificial Intelligence; RPTU – U...
Poster
main
https://openaccess.thecvf.com/content/WACV2025/html/Hegde_Modality-Incremental_Learning_with_Disjoint_Relevance_Mapping_Networks_for_Image-Based_Semantic_WACV_2025_paper.html
1
2411.17610
Modality-Incremental Learning with Disjoint Relevance Mapping Networks for Image-Based Semantic Segmentation In autonomous driving environment perception has significantly advanced with the utilization of deep learning techniques for diverse sensors such as cameras depth sensors or infrared sensors. The diversity in th...
[ -0.05613592639565468, -0.02635064721107483, 0.013323049992322922, -0.004030153155326843, -0.04867575690150261, -0.026627633720636368, 0.04206501320004463, 0.0008459629025310278, -0.00509655149653554, -0.028271086513996124, -0.03059777431190014, -0.03425399586558342, -0.03453098237514496, 0...
wacv_2025_f02d0d5af3
f02d0d5af3
wacv
2,025
Moment of Untruth: Dealing with Negative Queries in Video Moment Retrieval
Video Moment Retrieval is a common task to evaluate the performance of visual-language models - it involves localising start and end times of moments in videos from query sentences. The current task formulation assumes that the queried moment is present in the video resulting in false positive moment predictions when i...
Kevin Flanagan; Dima Damen; Michael Wray
University of Bristol; University of Bristol; University of Bristol
Poster
main
https://github.com/keflanagan/MomentofUntruth
https://openaccess.thecvf.com/content/WACV2025/html/Flanagan_Moment_of_Untruth_Dealing_with_Negative_Queries_in_Video_Moment_WACV_2025_paper.html
0
2502.08544
Moment of Untruth: Dealing with Negative Queries in Video Moment Retrieval Video Moment Retrieval is a common task to evaluate the performance of visual-language models - it involves localising start and end times of moments in videos from query sentences. The current task formulation assumes that the queried moment is...
[ -0.015858177095651627, -0.048933807760477066, -0.000710786203853786, 0.020824052393436432, -0.015468520112335682, 0.028997810557484627, 0.02834535948932171, 0.013828330673277378, -0.007607394363731146, 0.0031580429058521986, -0.017706789076328278, -0.043822940438985825, -0.014852316118776798...
wacv_2025_03447a5bad
03447a5bad
wacv
2,025
MonoPP: Metric-Scaled Self-Supervised Monocular Depth Estimation by Planar-Parallax Geometry in Automotive Applications
Self-supervised monocular depth estimation (MDE) has gained popularity for obtaining depth predictions directly from videos. However these methods often produce scale-invariant results unless additional training signals are provided. Addressing this challenge we introduce a novel self-supervised metric-scaled MDE model...
Gasser Elazab; Torben Gräber; Michael Unterreiner; Olaf Hellwich
;;;
Poster
main
https://openaccess.thecvf.com/content/WACV2025/html/Elazab_MonoPP_Metric-Scaled_Self-Supervised_Monocular_Depth_Estimation_by_Planar-Parallax_Geometry_in_WACV_2025_paper.html
0
MonoPP: Metric-Scaled Self-Supervised Monocular Depth Estimation by Planar-Parallax Geometry in Automotive Applications Self-supervised monocular depth estimation (MDE) has gained popularity for obtaining depth predictions directly from videos. However these methods often produce scale-invariant results unless addition...
[ -0.05358601734042168, -0.05885126441717148, -0.032618388533592224, -0.0030853969510644674, 0.002252199687063694, 0.026064835488796234, 0.015076901763677597, 0.01898849382996559, -0.010437137447297573, 0.006754265166819096, -0.052727147936820984, -0.030191145837306976, 0.007020328193902969, ...
wacv_2025_3f98b06f9c
3f98b06f9c
wacv
2,025
MulModSeg: Enhancing Unpaired Multi-Modal Medical Image Segmentation with Modality-Conditioned Text Embedding and Alternating Training
In the diverse field of medical imaging automatic segmentation has numerous applications and must handle a wide variety of input domains such as different types of Computed Tomography (CT) scans and Magnetic Resonance (MR) images. This heterogeneity challenges automatic segmentation algorithms to maintain consistent pe...
Chengyin Li; Hui Zhu; Rafi Ibn Sultan; Hassan Bagher Ebadian; Prashant Khanduri; Chetty Indrin; Kundan Thind; Dongxiao Zhu
Wayne State University+Henry Ford Health; Wayne State University; Wayne State University; Henry Ford Health; Wayne State University; Cedars Sinai Medical Center; Henry Ford Health; Wayne State University
Poster
main
https://openaccess.thecvf.com/content/WACV2025/html/Li_MulModSeg_Enhancing_Unpaired_Multi-Modal_Medical_Image_Segmentation_with_Modality-Conditioned_Text_WACV_2025_paper.html
3
2411.15576
MulModSeg: Enhancing Unpaired Multi-Modal Medical Image Segmentation with Modality-Conditioned Text Embedding and Alternating Training In the diverse field of medical imaging automatic segmentation has numerous applications and must handle a wide variety of input domains such as different types of Computed Tomography (...
[ -0.06310158222913742, -0.030443746596574783, -0.02656908705830574, 0.019354842603206635, 0.012601865455508232, 0.0005647661164402962, 0.009409884922206402, 0.009686646983027458, 0.006056460551917553, 0.008076817728579044, -0.026089368388056755, -0.05428212136030197, 0.013164614327251911, 0...
wacv_2025_f5a5c16c34
f5a5c16c34
wacv
2,025
Multi-Aperture Transformers for 3D (MAT3D) Segmentation of Clinical and Microscopic Images
3D segmentation of biological structures is critical in biomedical imaging offering significant insights into structures and functions. This paper introduces a novel segmentation of biological images that couples Multi-Aperture representation with Transformers for 3D (MAT3D) segmentation. Our method integrates the glob...
Muhammad Sohaib; Siyavash Shabani; Sahar A. Mohammed; Garrett Winkelmaier; Bahram Parvin
1Department of Electrical and Biomedical Engineering, University of Nevada, Reno, USA+2Pennington Cancer Institute; 1Department of Electrical and Biomedical Engineering, University of Nevada, Reno, USA+2Pennington Cancer Institute; 1Department of Electrical and Biomedical Engineering, University of Nevada, Reno, USA; 1...
Poster
main
https://github.com/sohaibcs1/MAT3D
https://openaccess.thecvf.com/content/WACV2025/html/Sohaib_Multi-Aperture_Transformers_for_3D_MAT3D_Segmentation_of_Clinical_and_Microscopic_WACV_2025_paper.html
0
Multi-Aperture Transformers for 3D (MAT3D) Segmentation of Clinical and Microscopic Images 3D segmentation of biological structures is critical in biomedical imaging offering significant insights into structures and functions. This paper introduces a novel segmentation of biological images that couples Multi-Aperture r...
[ -0.031968507915735245, -0.056559666991233826, -0.03525344654917717, 0.02789444848895073, 0.012800249271094799, 0.0014738637255504727, 0.03952937200665474, -0.013433379121124744, -0.004340155981481075, 0.04430079087615013, -0.020718969404697418, 0.001489921472966671, 0.0026976868975907564, ...
wacv_2025_b7539ea015
b7539ea015
wacv
2,025
Multi-Class Textual-Inversion Secretly Yields a Semantic-Agnostic Classifier
With the advent of large pre-trained vision-language models such as CLIP prompt learning methods aim to enhance the transferability of the CLIP model. They learn the prompt given few samples from the downstream task given the specific class names as prior knowledge which we term as semantic-aware classification. Howeve...
Kai Wang; Fei Yang; Bogdan Raducanu; Joost van de Weijer
Computer Vision Center + Universitat Autònoma de Barcelona, Spain; VCIP, College of Computer Science, Nankai University, China + Nankai International Advanced Research Institute (SHENZHEN· FUTIAN), China; Computer Vision Center + Universitat Autònoma de Barcelona, Spain; Computer Vision Center + Universitat Autònoma de...
Poster
main
https://openaccess.thecvf.com/content/WACV2025/html/Wang_Multi-Class_Textual-Inversion_Secretly_Yields_a_Semantic-Agnostic_Classifier_WACV_2025_paper.html
2
2410.22317
Multi-Class Textual-Inversion Secretly Yields a Semantic-Agnostic Classifier With the advent of large pre-trained vision-language models such as CLIP prompt learning methods aim to enhance the transferability of the CLIP model. They learn the prompt given few samples from the downstream task given the specific class na...
[ -0.0833628922700882, -0.019134897738695145, 0.029351303353905678, 0.038010213524103165, 0.009752864949405193, -0.039456456899642944, -0.052954718470573425, 0.026199236512184143, 0.016529807820916176, -0.0041788057424128056, -0.022398214787244797, -0.03333773836493492, -0.02180488407611847, ...
wacv_2025_6a34fd8e17
6a34fd8e17
wacv
2,025
Multi-HexPlanes: A Lightweight Map Representation for Rendering and 3D Reconstruction
Creating maps of the world around us is paramount to many applications including those related to robotics such as navigation and inspection. Given the computational resource limitations typical of robotic platforms there is a pressing need for lightweight 3D representations that capture detailed texture and geometric ...
Jianhao Zheng; Gábor Valasek; Daniel Barath; Iro Armeni
Stanford University; ELTE; ETH Zurich; Stanford University
Poster
main
https://openaccess.thecvf.com/content/WACV2025/html/Zheng_Multi-HexPlanes_A_Lightweight_Map_Representation_for_Rendering_and_3D_Reconstruction_WACV_2025_paper.html
0
Multi-HexPlanes: A Lightweight Map Representation for Rendering and 3D Reconstruction Creating maps of the world around us is paramount to many applications including those related to robotics such as navigation and inspection. Given the computational resource limitations typical of robotic platforms there is a pressin...
[ -0.03201154246926308, -0.02032742276787758, -0.005150964949280024, 0.004699449520558119, -0.030389772728085518, -0.021488461643457413, 0.010983810760080814, 0.023976406082510948, 0.03206682950258255, 0.0039208149537444115, -0.020935585722327232, 0.007919954136013985, 0.04658905044198036, 0...
wacv_2025_9267556795
9267556795
wacv
2,025
Multi-Label Continual Learning for the Medical Domain: A Novel Benchmark
Despite the critical importance of the medical domain in Deep Learning most of the research in this area solely focuses on training models in static environments. It is only in recent years that research has begun to address dynamic environments and tackle the Catastrophic Forgetting problem through Continual Learning ...
Marina Ceccon; Davide Dalle Pezze; Alessandro Fabris; Gian Antonio Susto
University of Padova, Padova, Italy; University of Padova, Padova, Italy; Max Planck Institute for Security and Privacy, Bochum, Germany; University of Padova, Padova, Italy
Poster
main
https://openaccess.thecvf.com/content/WACV2025/html/Ceccon_Multi-Label_Continual_Learning_for_the_Medical_Domain_A_Novel_Benchmark_WACV_2025_paper.html
4
2404.06859
Multi-Label Continual Learning for the Medical Domain: A Novel Benchmark Despite the critical importance of the medical domain in Deep Learning most of the research in this area solely focuses on training models in static environments. It is only in recent years that research has begun to address dynamic environments a...
[ -0.11479595303535461, -0.04070570692420006, -0.011713872663676739, 0.008675587363541126, -0.030401160940527916, 0.024050043895840645, 0.006502114702016115, 0.0030016798991709948, 0.027271360158920288, -0.0014402115484699607, -0.006177237723022699, -0.0052117579616606236, -0.02745438925921917...
wacv_2025_62678d429d
62678d429d
wacv
2,025
Multi-Level Feature Distillation of Joint Teachers Trained on Distinct Image Datasets
We propose a novel teacher-student framework to distill knowledge from multiple teachers trained on distinct datasets. Each teacher is first trained from scratch on its own dataset. Then the teachers are combined into a joint architecture which fuses the features of all teachers at multiple representation levels. The j...
Adrian Iordache; Bogdan Alexe; Radu Tudor Ionescu
;;
Poster
main
https://openaccess.thecvf.com/content/WACV2025/html/Iordache_Multi-Level_Feature_Distillation_of_Joint_Teachers_Trained_on_Distinct_Image_WACV_2025_paper.html
1
2410.22184
Multi-Level Feature Distillation of Joint Teachers Trained on Distinct Image Datasets We propose a novel teacher-student framework to distill knowledge from multiple teachers trained on distinct datasets. Each teacher is first trained from scratch on its own dataset. Then the teachers are combined into a joint architec...
[ -0.025204580277204514, -0.056006088852882385, -0.03282671049237251, -0.016846012324094772, -0.004512982442975044, -0.02297685667872429, -0.004552105907350779, 0.012786399573087692, 0.011875058524310589, -0.007212487049400806, -0.0067798299714922905, -0.006881089881062508, -0.0257016774266958...
wacv_2025_51684cbb29
51684cbb29
wacv
2,025
Multi-Modal Large Language Model with RAG Strategies in Soccer Commentary Generation
As a globally celebrated sport soccer has seen its appeal greatly amplified by engaging and vivid commentary. Recently Multi-Modal Large Language Models (MLLMs) have attracted attention in generating soccer commentaries due to their remarkable capacities of understanding different modalities of the input videos. Most o...
Xiang Li; Yangfan He; Shuaishuai Zu; Zhengyang Li; Tianyu Shi; Yiting Xie; Kevin Zhang
Southern University of Science and Technology; University of Minnesota-Twin Cities; KNQ.AI+Renmin University of China; KNQ.AI+Imperial College London; University of Toronto; KNQ.AI; KNQ.AI
Poster
main
https://openaccess.thecvf.com/content/WACV2025/html/Li_Multi-Modal_Large_Language_Model_with_RAG_Strategies_in_Soccer_Commentary_WACV_2025_paper.html
0
Multi-Modal Large Language Model with RAG Strategies in Soccer Commentary Generation As a globally celebrated sport soccer has seen its appeal greatly amplified by engaging and vivid commentary. Recently Multi-Modal Large Language Models (MLLMs) have attracted attention in generating soccer commentaries due to their re...
[ -0.027672115713357925, -0.02140568383038044, -0.012172403745353222, 0.009575254283845425, -0.005684151779860258, 0.021461140364408493, -0.010416324250400066, 0.009062293916940689, 0.005208162125200033, -0.04562109708786011, -0.031110333278775215, -0.018327923491597176, -0.025601789355278015,...
wacv_2025_3a9a866ae9
3a9a866ae9
wacv
2,025
Multi-Modal Large Language Models are Effective Vision Learners
Large language models (LLMs) pre-trained on vast amounts of text have shown remarkable abilities in understanding general knowledge and commonsense. Therefore it's desirable to leverage pre-trained LLM to help solve computer vision tasks. Previous works on multi-modal LLM mainly focus on the generation capability. In t...
Li Sun; Chaitanya Ahuja; Peng Chen; Matt D'Zmura; Kayhan Batmanghelich; Philip Bontrager
Boston University; Meta; Meta; Meta; Boston University; Meta
Poster
main
https://openaccess.thecvf.com/content/WACV2025/html/Sun_Multi-Modal_Large_Language_Models_are_Effective_Vision_Learners_WACV_2025_paper.html
0
Multi-Modal Large Language Models are Effective Vision Learners Large language models (LLMs) pre-trained on vast amounts of text have shown remarkable abilities in understanding general knowledge and commonsense. Therefore it's desirable to leverage pre-trained LLM to help solve computer vision tasks. Previous works on...
[ -0.03146059438586235, -0.0033784920815378428, 0.019684001803398132, -0.015279518440365791, -0.03841009736061096, -0.007052808068692684, -0.012950495816767216, -0.01901722513139248, 0.01101590320467949, -0.03157328814268112, -0.025976121425628662, -0.03519830107688904, -0.020886078476905823, ...
wacv_2025_a6006328f1
a6006328f1
wacv
2,025
Multi-Resolution Guided 3D GANs for Medical Image Translation
Medical image translation is the process of converting from one imaging modality to another in order to reduce the need for multiple image acquisitions from the same patient. This can enhance the efficiency of treatment by reducing the time equipment and labor needed. In this paper we introduce a multi-resolution guide...
Juhyung Ha; Jong Sung Park; David Crandall; Eleftherios Garyfallidis; Xuhong Zhang
Indiana University Bloomington; Indiana University Bloomington; Indiana University Bloomington; Indiana University Bloomington; Indiana University Bloomington
Poster
main
github.com/juhha/3D-mADUNet
https://openaccess.thecvf.com/content/WACV2025/html/Ha_Multi-Resolution_Guided_3D_GANs_for_Medical_Image_Translation_WACV_2025_paper.html
0
2412.00575
Multi-Resolution Guided 3D GANs for Medical Image Translation Medical image translation is the process of converting from one imaging modality to another in order to reduce the need for multiple image acquisitions from the same patient. This can enhance the efficiency of treatment by reducing the time equipment and lab...
[ -0.0267924927175045, -0.039572715759277344, -0.04920845106244087, -0.036465004086494446, -0.007562397047877312, 0.0034754846710711718, -0.00276291836053133, 0.03552717715501785, 0.024751335382461548, 0.01899563893675804, -0.047112125903367996, 0.020264465361833572, -0.014563935808837414, 0...
wacv_2025_3f621a27d8
3f621a27d8
wacv
2,025
Multi-Scale Grouped Prototypes for Interpretable Semantic Segmentation
Prototypical part learning is emerging as a promising approach for making semantic segmentation interpretable. The model selects real patches seen during training as prototypes and constructs the dense prediction map based on the similarity between parts of the test image and the prototypes. This improves interpretabil...
Hugo Porta; Emanuele Dalsasso; Diego Marcos; Devis Tuia
;;;
Poster
main
github.com/eceo-epfl/ScaleProtoSeg
https://openaccess.thecvf.com/content/WACV2025/html/Porta_Multi-Scale_Grouped_Prototypes_for_Interpretable_Semantic_Segmentation_WACV_2025_paper.html
0
2409.09497
Multi-Scale Grouped Prototypes for Interpretable Semantic Segmentation Prototypical part learning is emerging as a promising approach for making semantic segmentation interpretable. The model selects real patches seen during training as prototypes and constructs the dense prediction map based on the similarity between ...
[ -0.020855246111750603, -0.011837249621748924, 0.0023629318457096815, 0.023963652551174164, 0.0043327948078513145, -0.04004785418510437, 0.03762618824839592, -0.019879350438714027, 0.03243948519229889, 0.04803573712706566, -0.03972255438566208, -0.0033456042874604464, -0.009894494898617268, ...
wacv_2025_3472e927a0
3472e927a0
wacv
2,025
Multi-Spectral Image Color Reproduction
From camera to screen researchers have developed a well-established system for capturing and reproducing the color experience of human eyes. In this study we aim to upgrade this process by transiting from conventional RGB to multi-spectral image (MSI) color reproduction. While MSI offers evident advantages in color mat...
Jiacheng Li; Chang Chen; Xue Hu; Fenglong Song; Youliang Yan; Zhiwei Xiong
University of Science and Technology of China; Huawei Noah’s Ark Lab; Huawei Noah’s Ark Lab; Huawei Noah’s Ark Lab; Huawei Noah’s Ark Lab; University of Science and Technology of China
Poster
main
https://openaccess.thecvf.com/content/WACV2025/html/Li_Multi-Spectral_Image_Color_Reproduction_WACV_2025_paper.html
0
Multi-Spectral Image Color Reproduction From camera to screen researchers have developed a well-established system for capturing and reproducing the color experience of human eyes. In this study we aim to upgrade this process by transiting from conventional RGB to multi-spectral image (MSI) color reproduction. While MS...
[ -0.03175823390483856, -0.018363477662205696, -0.011023754253983498, 0.029283322393894196, -0.01296967826783657, -0.02382339909672737, -0.002831508405506611, 0.010740366764366627, 0.0577731616795063, 0.025788215920329094, -0.04772236943244934, -0.015208435244858265, -0.0012209255946800113, ...
wacv_2025_13f5c74daa
13f5c74daa
wacv
2,025
Multi-Surrogate-Teacher Assistance for Representation Alignment in Fingerprint-Based Indoor Localization
Despite remarkable progress in knowledge transfer across visual and textual domains extending these achievements to indoor localization particularly for learning transferable representations among Received Signal Strength (RSS) fingerprint datasets remains a challenge. This is due to inherent discrepancies among these ...
Son Minh Nguyen; Linh Duy Tran; Duc Le; Paul Havinga
Department of Computer Science, University of Twente; Viettel AI, Viettel Group; Department of Computer Science, University of Twente; Department of Computer Science, University of Twente
Poster
main
https://github.com/Minh-Son-Nguyen/RSS_TL
https://openaccess.thecvf.com/content/WACV2025/html/Nguyen_Multi-Surrogate-Teacher_Assistance_for_Representation_Alignment_in_Fingerprint-Based_Indoor_Localization_WACV_2025_paper.html
0
2412.12189
Multi-Surrogate-Teacher Assistance for Representation Alignment in Fingerprint-Based Indoor Localization Despite remarkable progress in knowledge transfer across visual and textual domains extending these achievements to indoor localization particularly for learning transferable representations among Received Signal St...
[ -0.05785717815160751, 0.004839904606342316, -0.03338056430220604, -0.00251000770367682, -0.016080675646662712, -0.0017041267128661275, -0.018380552530288696, 0.011277716606855392, 0.012607766315340996, 0.03979066386818886, 0.012053579092025757, 0.02682267688214779, 0.013873161748051643, 0....
wacv_2025_912e6802df
912e6802df
wacv
2,025
Multi-Task Learning of Classification and Generation for Set-Structured Data
In this study we propose a multi-task learning model of classification and generation for set-structured data. The proposed model learns data generation and classification in a single neural network by integrating a classification layer into a variational autoencoder while maintaining permutation invariance and equivar...
Fumioki Sato; Hideaki Hayashi; Hajime Nagahara
Osaka University; Osaka University; Osaka University
Poster
main
https://openaccess.thecvf.com/content/WACV2025/html/Sato_Multi-Task_Learning_of_Classification_and_Generation_for_Set-Structured_Data_WACV_2025_paper.html
0
Multi-Task Learning of Classification and Generation for Set-Structured Data In this study we propose a multi-task learning model of classification and generation for set-structured data. The proposed model learns data generation and classification in a single neural network by integrating a classification layer into a...
[ -0.05208715423941612, -0.021223407238721848, -0.03991760313510895, 0.039954256266355515, 0.021223407238721848, -0.015532675199210644, -0.00246048578992486, 0.037021834403276443, 0.0005569884669966996, 0.00965866632759571, -0.041823677718639374, -0.032678183168172836, -0.030295589938759804, ...
wacv_2025_dc2512006b
dc2512006b
wacv
2,025
Multi-View Factorizing and Disentangling: A Novel Framework for Incomplete Multi-View Multi-Label Classification
Multi-view multi-label classification (MvMLC) has recently garnered significant research attention due to its wide range of real-world applications. However incompleteness in views and labels is a common challenge often resulting from data collection oversights and uncertainties in manual annotation. Furthermore the ta...
Wulin Xie; Lian Zhao; Jiang Long; Xiaohuan Lu; Bingyan Nie
Guizhou University; Guizhou University; Guizhou University; Guizhou University; Guizhou University
Poster
main
https://openaccess.thecvf.com/content/WACV2025/html/Xie_Multi-View_Factorizing_and_Disentangling_A_Novel_Framework_for_Incomplete_Multi-View_WACV_2025_paper.html
5
2501.06524
Multi-View Factorizing and Disentangling: A Novel Framework for Incomplete Multi-View Multi-Label Classification Multi-view multi-label classification (MvMLC) has recently garnered significant research attention due to its wide range of real-world applications. However incompleteness in views and labels is a common cha...
[ -0.038139209151268005, -0.022835759446024895, -0.03143361583352089, 0.013980701565742493, -0.012382382526993752, -0.015964822843670845, 0.01201495248824358, 0.0036283687222748995, 0.0291555505245924, 0.0005055029760114849, -0.018564388155937195, -0.047324951738119125, -0.008597856387495995, ...
wacv_2025_3511438eec
3511438eec
wacv
2,025
Multi-View Image Diffusion via Coordinate Noise and Fourier Attention
Recently text-to-image generation with diffusion models has made significant advancements in both higher fidelity and generalization capabilities compared to previous baselines. However generating holistic multi-view consistent images from prompts still remains an important and challenging task. To address this challen...
Justin Theiss; Norman Müller; Daeil Kim; Aayush Prakash
;;;
Poster
main
https://openaccess.thecvf.com/content/WACV2025/html/Theiss_Multi-View_Image_Diffusion_via_Coordinate_Noise_and_Fourier_Attention_WACV_2025_paper.html
0
Multi-View Image Diffusion via Coordinate Noise and Fourier Attention Recently text-to-image generation with diffusion models has made significant advancements in both higher fidelity and generalization capabilities compared to previous baselines. However generating holistic multi-view consistent images from prompts st...
[ -0.01564478687942028, -0.020288050174713135, -0.02835218422114849, 0.049625445157289505, -0.005569179076701403, 0.01597318984568119, -0.040466632694005966, 0.04999033734202385, 0.037729933857917786, -0.0013022118946537375, -0.024794477969408035, -0.011083624325692654, -0.022769322618842125, ...
wacv_2025_38700e9a19
38700e9a19
wacv
2,025
Multimodal Fusion Learning with Dual Attention for Medical Imaging
Multimodal fusion learning has shown significant promise in classifying various diseases such as skin cancer and brain tumors. However existing methods face three key limitations. First they often lack generalizability to other diagnosis tasks due to their focus on a particular disease. Second they do not fully leverag...
Joy Dhar; Nayyar Zaidi; Maryam Haghighat; Sudipta Roy; Puneet Goyal; Azadeh Alavi; Vikas Kumar
;;;;;;
Poster
main
https://openaccess.thecvf.com/content/WACV2025/html/Dhar_Multimodal_Fusion_Learning_with_Dual_Attention_for_Medical_Imaging_WACV_2025_paper.html
4
2412.01248
Multimodal Fusion Learning with Dual Attention for Medical Imaging Multimodal fusion learning has shown significant promise in classifying various diseases such as skin cancer and brain tumors. However existing methods face three key limitations. First they often lack generalizability to other diagnosis tasks due to th...
[ -0.02555961161851883, -0.02499202825129032, -0.029331302270293236, -0.008783828467130661, -0.032260771840810776, 0.0019167420687153935, -0.014336085878312588, 0.030393235385417938, 0.04566308856010437, 0.002416811417788267, -0.03251710161566734, 0.0032544564455747604, -0.03797323629260063, ...
wacv_2025_ed7113ad68
ed7113ad68
wacv
2,025
Multimodal Interpretable Depression Analysis using Visual Physiological Audio and Textual Data
Motivated by depression's significant impact on global health this work proposes MultiDepNet a novel multimodal interpretable depression detection system integrating visual physiological audio and textual data. Through dedicated feature extraction methods (MTCNN for video TS-CAN for physiological ResNet-18 for audio an...
Puneet Kumar; Shreshtha Misra; Zhuhong Shao; Bin Zhu; Balasubramanian Raman; Xiaobai Li
University of Oulu, Finland; IIT Roorkee, India; Capital Normal Univ., China; Hangzhou City Univ., China; IIT Roorkee, India; Zhejiang University, China
Poster
main
https://openaccess.thecvf.com/content/WACV2025/html/Kumar_Multimodal_Interpretable_Depression_Analysis_using_Visual_Physiological_Audio_and_Textual_WACV_2025_paper.html
0
Multimodal Interpretable Depression Analysis using Visual Physiological Audio and Textual Data Motivated by depression's significant impact on global health this work proposes MultiDepNet a novel multimodal interpretable depression detection system integrating visual physiological audio and textual data. Through dedica...
[ -0.041331708431243896, -0.04323464259505272, -0.0525590144097805, 0.005913365166634321, -0.01437666080892086, 0.003767807735130191, 0.01343470811843872, 0.010209236294031143, 0.04688827320933342, 0.030408872291445732, -0.06055133417248726, -0.01856311410665512, -0.04072277247905731, 0.0430...
wacv_2025_232baa3b91
232baa3b91
wacv
2,025
Multispectral Object Detection Enhanced by Cross-Modal Information Complementary and Cosine Similarity Channel Resampling Modules
Images obtained from different modalities can effectively enhance the accuracy and reliability of the detection model by complementing specialized information from visible (RGB) and infrared (IR) images. However integrating information from multiple modalities faces the following challenges: 1) distinct characteristics...
Junbo Jang; Chanyeong Park; Heegwang Kim; Jiyoon Lee; Joonki Paik
Chung-Ang University, Republic of Korea; Chung-Ang University, Republic of Korea; Chung-Ang University, Republic of Korea; Chung-Ang University, Republic of Korea; Chung-Ang University, Republic of Korea
Poster
main
https://openaccess.thecvf.com/content/WACV2025/html/Jang_Multispectral_Object_Detection_Enhanced_by_Cross-Modal_Information_Complementary_and_Cosine_WACV_2025_paper.html
0
Multispectral Object Detection Enhanced by Cross-Modal Information Complementary and Cosine Similarity Channel Resampling Modules Images obtained from different modalities can effectively enhance the accuracy and reliability of the detection model by complementing specialized information from visible (RGB) and infrared...
[ -0.06023421138525009, -0.05341454967856407, -0.02501157484948635, 0.02403470315039158, -0.006741329096257687, -0.009805569425225258, -0.009455370716750622, 0.018486816436052322, 0.016376405954360962, 0.017666613683104515, -0.04814313352108002, -0.05625300481915474, -0.0068380944430828094, ...
wacv_2025_b755a79805
b755a79805
wacv
2,025
My3DGen: A Scalable Personalized 3D Generative Model
In recent years generative 3D face models (e.g. EG3D) have been developed to tackle the problem of synthesizing photo-realistic faces. However these models are often unable to capture facial features unique to each individual highlighting the importance of personalization. Some prior works have shown promise in persona...
Luchao Qi; Jiaye Wu; Annie N. Wang; Shengze Wang; Roni Sengupta
The University of North Carolina at Chapel Hill, USA; The University of Maryland, USA; The University of North Carolina at Chapel Hill, USA; The University of North Carolina at Chapel Hill, USA; The University of North Carolina at Chapel Hill, USA
Poster
main
https://openaccess.thecvf.com/content/WACV2025/html/Qi_My3DGen_A_Scalable_Personalized_3D_Generative_Model_WACV_2025_paper.html
2
2307.05468
My3DGen: A Scalable Personalized 3D Generative Model In recent years generative 3D face models (e.g. EG3D) have been developed to tackle the problem of synthesizing photo-realistic faces. However these models are often unable to capture facial features unique to each individual highlighting the importance of personaliz...
[ -0.010749256238341331, -0.026813166216015816, -0.02166459523141384, -0.05373705178499222, -0.0022605732083320618, -0.009074586443603039, 0.0012686890549957752, -0.012197867967188358, 0.011404360644519329, -0.019468609243631363, -0.0432184673845768, 0.020391292870044708, -0.01162580493837595,...
wacv_2025_39e9ead91e
39e9ead91e
wacv
2,025
NAT: Learning to Attack Neurons for Enhanced Adversarial Transferability
The generation of transferable adversarial perturbations typically involves training a generator to maximize embedding separation between clean and adversarial images at a single mid-layer of a source model. In this work we build on this approach and introduce Neuron Attack for Transferability (NAT) a method designed t...
Krishna Kanth Nakka; Alexandre Alahi
VITA Lab, EPFL, Switzerland; VITA Lab, EPFL, Switzerland
Poster
main
https://krishnakanthnakka.github.io/NAT/
https://openaccess.thecvf.com/content/WACV2025/html/Nakka_NAT_Learning_to_Attack_Neurons_for_Enhanced_Adversarial_Transferability_WACV_2025_paper.html
0
NAT: Learning to Attack Neurons for Enhanced Adversarial Transferability The generation of transferable adversarial perturbations typically involves training a generator to maximize embedding separation between clean and adversarial images at a single mid-layer of a source model. In this work we build on this approach ...
[ -0.0014664947520941496, -0.022145314142107964, -0.018586890771985054, 0.02716684341430664, -0.020501574501395226, -0.029515042901039124, 0.023066530004143715, -0.007076204288750887, -0.016374165192246437, 0.011172003112733364, -0.0257579255849123, 0.009162487462162971, -0.03520490601658821, ...
wacv_2025_3ee776bd65
3ee776bd65
wacv
2,025
NCAP: Scene Text Image Super-Resolution with Non-CAtegorical Prior
Scene text image super-resolution (STISR) enhances the resolution and quality of low-resolution images. Unlike previous studies that treated scene text images as natural images recent methods using a text prior (TP) extracted from a pre-trained text recognizer have shown strong performance. However two major issues eme...
Dongwoo Park; Suk Pil Ko
THINKWARE Corporation, Republic of Korea; THINKWARE Corporation, Republic of Korea
Poster
main
https://openaccess.thecvf.com/content/WACV2025/html/Park_NCAP_Scene_Text_Image_Super-Resolution_with_Non-CAtegorical_Prior_WACV_2025_paper.html
0
NCAP: Scene Text Image Super-Resolution with Non-CAtegorical Prior Scene text image super-resolution (STISR) enhances the resolution and quality of low-resolution images. Unlike previous studies that treated scene text images as natural images recent methods using a text prior (TP) extracted from a pre-trained text rec...
[ -0.05948168411850929, -0.026815207675099373, -0.002066112821921706, 0.0057674129493534565, -0.008818830363452435, -0.006885489448904991, -0.02353551611304283, 0.042635995894670486, -0.00855328794568777, 0.051766954362392426, -0.028268707916140556, -0.04010168835520744, -0.031157072633504868,...
wacv_2025_9d381d1e8e
9d381d1e8e
wacv
2,025
NCAdapt: Dynamic Adaptation with Domain-Specific Neural Cellular Automata for Continual Hippocampus Segmentation
Continual learning (CL) in medical imaging presents a unique challenge requiring models to adapt to new domains while retaining previously acquired knowledge. We introduce NCAdapt a Neural Cellular Automata (NCA) based method designed to address this challenge. NCAdapt features a domain-specific multi-head structure in...
Amin Ranem; John Orlando Kalkhof; Anirban Mukhopadhyay
Technical University of Darmstadt; Technical University of Darmstadt; Technical University of Darmstadt
Poster
main
https://openaccess.thecvf.com/content/WACV2025/html/Ranem_NCAdapt_Dynamic_Adaptation_with_Domain-Specific_Neural_Cellular_Automata_for_Continual_WACV_2025_paper.html
0
2410.23368
NCAdapt: Dynamic Adaptation with Domain-Specific Neural Cellular Automata for Continual Hippocampus Segmentation Continual learning (CL) in medical imaging presents a unique challenge requiring models to adapt to new domains while retaining previously acquired knowledge. We introduce NCAdapt a Neural Cellular Automata ...
[ -0.0379432775080204, -0.017677277326583862, -0.010003563947975636, 0.02220754325389862, -0.019896183162927628, -0.03934858739376068, 0.027884243056178093, 0.009633746929466724, 0.044489048421382904, 0.017251988872885704, -0.006291520781815052, -0.013331921771168709, -0.012906632386147976, ...
wacv_2025_7e8c46c7f5
7e8c46c7f5
wacv
2,025
NPL-MVPS: Neural Point-Light Multi-View Photometric Stereo
In this work we present a novel multi-view photometric stereo (MVPS) method. Like many works in 3D reconstruction we are leveraging neural shape representations and learnt renderers. However our work differs from the state-of-the-art multi-view PS methods such as PS-NeRF or Supernormal in that we explicitly leverage pe...
Fotios Logothetis; Ignas Budvytis; Roberto Cipolla
Toshiba Europe; Independent researcher; University of Cambridge
Poster
main
https://openaccess.thecvf.com/content/WACV2025/html/Logothetis_NPL-MVPS_Neural_Point-Light_Multi-View_Photometric_Stereo_WACV_2025_paper.html
2
NPL-MVPS: Neural Point-Light Multi-View Photometric Stereo In this work we present a novel multi-view photometric stereo (MVPS) method. Like many works in 3D reconstruction we are leveraging neural shape representations and learnt renderers. However our work differs from the state-of-the-art multi-view PS methods such ...
[ -0.02394634112715721, -0.02152956835925579, -0.02252579666674137, 0.011973170563578606, -0.016723696142435074, 0.006600002758204937, 0.02597569301724434, 0.00021893353550694883, 0.059478431940078735, 0.0038280938751995564, -0.029222654178738594, 0.01441761665046215, 0.01689895987510681, 0....
wacv_2025_189f63a204
189f63a204
wacv
2,025
NarrAD: Automatic Generation of Audio Descriptions for Movies with Rich Narrative Context
Audio Description (AD) is a narration designed to enhance accessibility for visually impaired individuals by conveying the key visual elements of a video. Thus automating AD generation for long-form videos such as movies and dramas provides high social value but is a challenging task. First AD must reflect the narrativ...
Jaehyeong Park; Junchel Ye; Seungkook Lee; Hyun W. Ka; Dongsu Han
KAIST, Republic of Korea; KAIST, Republic of Korea; KAIST, Republic of Korea; KAIST, Republic of Korea; KAIST, Republic of Korea
Poster
main
https://openaccess.thecvf.com/content/WACV2025/html/Park_NarrAD_Automatic_Generation_of_Audio_Descriptions_for_Movies_with_Rich_WACV_2025_paper.html
0
NarrAD: Automatic Generation of Audio Descriptions for Movies with Rich Narrative Context Audio Description (AD) is a narration designed to enhance accessibility for visually impaired individuals by conveying the key visual elements of a video. Thus automating AD generation for long-form videos such as movies and drama...
[ -0.04835640639066696, -0.033065326511859894, -0.021172264590859413, 0.004142500925809145, -0.01333068497478962, 0.017540866509079933, 0.030731523409485817, 0.021900411695241928, -0.014609609730541706, 0.020182732492685318, -0.0011721528135240078, -0.06232188642024994, -0.01029674056917429, ...
wacv_2025_e007a699ae
e007a699ae
wacv
2,025
Navigating Heterogeneity and Privacy in One-Shot Federated Learning with Diffusion Models
Federated learning (FL) enables multiple clients to train models collectively while preserving data privacy. However FL faces challenges in terms of communication cost and data heterogeneity. One-shot federated learning has emerged as a solution by reducing communication rounds improving efficiency and providing better...
Matias Mendieta; Guangyu Sun; Chen Chen
Center for Research in Computer Vision, University of Central Florida, USA; Center for Research in Computer Vision, University of Central Florida, USA; Center for Research in Computer Vision, University of Central Florida, USA
Poster
main
https://github.com/mmendiet/FedDiff
https://openaccess.thecvf.com/content/WACV2025/html/Mendieta_Navigating_Heterogeneity_and_Privacy_in_One-Shot_Federated_Learning_with_Diffusion_WACV_2025_paper.html
3
2405.01494
Navigating Heterogeneity and Privacy in One-Shot Federated Learning with Diffusion Models Federated learning (FL) enables multiple clients to train models collectively while preserving data privacy. However FL faces challenges in terms of communication cost and data heterogeneity. One-shot federated learning has emerge...
[ -0.054649293422698975, -0.03642673417925835, -0.05987149849534035, -0.02489742450416088, -0.008384952321648598, -0.006789788603782654, 0.007042624056339264, 0.014673666097223759, -0.0048636398278176785, 0.026147810742259026, -0.05365633964538574, -0.03216070309281349, -0.031848106533288956, ...
wacv_2025_62f5866e88
62f5866e88
wacv
2,025
NeRFs are Mirror Detectors: using Structural Similarity for Multi-View Mirror Scene Reconstruction with 3D Surface Primitives
While neural radiance fields (NeRF) led to a breakthrough in photorealistic novel view synthesis handling mirroring surfaces still denotes a particular challenge as they introduce severe inconsistencies in the scene representation. Previous attempts either focus on reconstructing single reflective objects or rely on st...
Leif Van Holland; Michael Weinmann; Jan U. Müller; Patrick Stotko; Reinhard Klein
University of Bonn; Delft University of Technology; University of Bonn; University of Bonn; University of Bonn
Poster
main
https://openaccess.thecvf.com/content/WACV2025/html/Van_Holland_NeRFs_are_Mirror_Detectors_using_Structural_Similarity_for_Multi-View_Mirror_WACV_2025_paper.html
0
NeRFs are Mirror Detectors: using Structural Similarity for Multi-View Mirror Scene Reconstruction with 3D Surface Primitives While neural radiance fields (NeRF) led to a breakthrough in photorealistic novel view synthesis handling mirroring surfaces still denotes a particular challenge as they introduce severe inconsi...
[ -0.024718645960092545, 0.015665579587221146, -0.006446741055697203, -0.011898841708898544, 0.003322373842820525, 0.014293345622718334, 0.030852260068058968, 0.02711315080523491, 0.053747400641441345, 0.035199206322431564, -0.034444015473127365, -0.022858301177620888, -0.014698568731546402, ...
wacv_2025_d62f1c2720
d62f1c2720
wacv
2,025
Needles & Haystacks: Dataset and Benchmark for Domain-Agnostic Image-Based Rigid Slice-to-Volume Registration
We address domain-agnostic slice-to-volume (S2V) registration the alignment of 2D sliced/tomographic images into 3D volumes without prior knowledge of structure shape or orientation. While S2V registration is well-studied in medical imaging which often relies on auxiliary information (e.g. landmarks segmentation masks ...
Anton Frolov; Florian Kleiner; Christiane Rößler; Volker Rodehorst
Bauhaus-Universität Weimar; Bauhaus-Universität Weimar; Bauhaus-Universität Weimar; Bauhaus-Universität Weimar
Poster
main
https://openaccess.thecvf.com/content/WACV2025/html/Frolov_Needles__Haystacks_Dataset_and_Benchmark_for_Domain-Agnostic_Image-Based_Rigid_WACV_2025_paper.html
0
Needles & Haystacks: Dataset and Benchmark for Domain-Agnostic Image-Based Rigid Slice-to-Volume Registration We address domain-agnostic slice-to-volume (S2V) registration the alignment of 2D sliced/tomographic images into 3D volumes without prior knowledge of structure shape or orientation. While S2V registration is w...
[ -0.04498636722564697, -0.00260631600394845, 0.009557269513607025, 0.02609582617878914, -0.00240331725217402, 0.02437850460410118, 0.00183282105717808, -0.030837127938866615, 0.0010208265157416463, 0.029175804927945137, -0.01366390846669674, 0.011927920393645763, 0.015502562746405602, 0.023...
wacv_2025_2d5658f057
2d5658f057
wacv
2,025
Negative-Prompt Inversion: Fast Image Inversion for Editing with Text-Guided Diffusion Models
In image editing employing diffusion models it is crucial to preserve the reconstruction fidelity to the original image while changing its style. Although existing methods ensure reconstruction fidelity through optimization a drawback of these is the significant amount of time required for optimization. In this paper w...
Daiki Miyake; Akihiro Iohara; Yu Saito; Toshiyuki Tanaka
The University of Tokyo, Japan + DATAGRID Inc., Japan; DATAGRID Inc., Japan; DATAGRID Inc., Japan; Kyoto University, Japan
Poster
main
https://openaccess.thecvf.com/content/WACV2025/html/Miyake_Negative-Prompt_Inversion_Fast_Image_Inversion_for_Editing_with_Text-Guided_Diffusion_WACV_2025_paper.html
106
2305.16807
Negative-Prompt Inversion: Fast Image Inversion for Editing with Text-Guided Diffusion Models In image editing employing diffusion models it is crucial to preserve the reconstruction fidelity to the original image while changing its style. Although existing methods ensure reconstruction fidelity through optimization a ...
[ -0.08381882309913635, -0.011632408015429974, 0.011214042082428932, 0.02717563323676586, 0.015934307128190994, -0.008458280004560947, 0.003794857067987323, 0.03388768807053566, 0.02830340526998043, 0.0436556376516819, -0.03392406925559044, -0.03277810662984848, 0.00037971307756379247, 0.002...
wacv_2025_b17a04f0c9
b17a04f0c9
wacv
2,025
NestedMorph: Enhancing Deformable Medical Image Registration with Nested Attention Mechanisms
Deformable image registration is crucial for aligning medical images in a nonlinear fashion across different modalities allowing for precise spatial correspondence between varying anatomical structures. This paper presents NestedMorph a novel network utilizing a Nested Attention Fusion approach to improve intra-subject...
Gurucharan Marthi Krishna Kumar; Janine Mendola; Amir Shmuel
Montreal Neurological Institute, McGill University; Dept. of Ophthalmology, McGill University; Montreal Neurological Institute, McGill University
Poster
main
https://github.com/AS-Lab/Marthi-et-al-2024-NestedMorph-Deformable-Medical-Image-Registration
https://openaccess.thecvf.com/content/WACV2025/html/Kumar_NestedMorph_Enhancing_Deformable_Medical_Image_Registration_with_Nested_Attention_Mechanisms_WACV_2025_paper.html
0
2410.02550
NestedMorph: Enhancing Deformable Medical Image Registration with Nested Attention Mechanisms Deformable image registration is crucial for aligning medical images in a nonlinear fashion across different modalities allowing for precise spatial correspondence between varying anatomical structures. This paper presents Nes...
[ -0.029417037963867188, -0.016053978353738785, 0.009967378340661526, 0.016401782631874084, -0.011056559160351753, 0.011029100976884365, 0.013280685059726238, -0.01681365817785263, 0.046020183712244034, 0.02551795355975628, -0.0281173437833786, -0.019550340250134468, 0.0005091234343126416, 0...
wacv_2025_ff5cb5009f
ff5cb5009f
wacv
2,025
NeuManifold: Neural Watertight Manifold Reconstruction with Efficient and High-Quality Rendering Support
While existing volumetric rendering approaches provide photorealistic results extracting high-quality meshes from optimized neural field representations is challenging. Conversely existing differentiable rasterization-based methods are typically sensitive to initialization and suffer from poor mesh rendering quality. I...
Xinyue Wei; Fanbo Xiang; Sai Bi; Anpei Chen; Kalyan Sunkavalli; Zexiang Xu; Hao Su
University of California San Diego; University of California San Diego; Adobe Research; ETH Zürich+University of Tübingen; Adobe Research; Adobe Research; University of California San Diego
Poster
main
https://sarahweiii.github.io/neumanifold/
https://openaccess.thecvf.com/content/WACV2025/html/Wei_NeuManifold_Neural_Watertight_Manifold_Reconstruction_with_Efficient_and_High-Quality_Rendering_WACV_2025_paper.html
15
2305.17134
NeuManifold: Neural Watertight Manifold Reconstruction with Efficient and High-Quality Rendering Support While existing volumetric rendering approaches provide photorealistic results extracting high-quality meshes from optimized neural field representations is challenging. Conversely existing differentiable rasterizati...
[ -0.02191799134016037, 0.0004220379050821066, 0.0023783352226018906, -0.009961027652025223, 0.0007461443310603499, -0.04357483237981796, 0.0181872695684433, 0.0227014422416687, 0.030181540176272392, 0.03850105032324791, -0.004845275077968836, -0.035609740763902664, -0.005759301595389843, 0....
wacv_2025_27998c5ea5
27998c5ea5
wacv
2,025
Neural Graph Map: Dense Mapping with Efficient Loop Closure Integration
Neural field-based SLAM methods typically employ a single monolithic field as their scene representation. This prevents efficient incorporation of loop closure constraints and limits scalability. To address these shortcomings we propose a novel RGB-D neural mapping framework in which the scene is represented by a colle...
Leonard Bruns; Jun Zhang; Patric Jensfelt
KTH Royal Institute of Technology, Stockholm, Sweden; TU Graz, Graz, Austria; KTH Royal Institute of Technology, Stockholm, Sweden
Poster
main
https://github.com/KTH-RPL/neural_graph_mapping
https://openaccess.thecvf.com/content/WACV2025/html/Bruns_Neural_Graph_Map_Dense_Mapping_with_Efficient_Loop_Closure_Integration_WACV_2025_paper.html
0
Neural Graph Map: Dense Mapping with Efficient Loop Closure Integration Neural field-based SLAM methods typically employ a single monolithic field as their scene representation. This prevents efficient incorporation of loop closure constraints and limits scalability. To address these shortcomings we propose a novel RGB...
[ -0.056211601942777634, -0.017481807619333267, -0.016788531094789505, -0.020592182874679565, -0.02096692845225334, -0.02904266119003296, 0.006525230128318071, -0.0011745882220566273, 0.019524162635207176, 0.024152250960469246, -0.002744999947026372, 0.015973463654518127, 0.032096825540065765,...
wacv_2025_c8a6b51715
c8a6b51715
wacv
2,025
Neural SDF for Shadow-Aware Unsupervised Structured Light
Among various active 3D measurement techniques Structured Light (SL) is one of the most popular method for its robustness and high accuracy. Ordinary SL system consists of a camera and a projector and by projecting a pre-defined pattern we can obtain pixel-to-pixel correspondences between the camera and the projector f...
Kazuto Ichimaru; Diego Thomas; Takafumi Iwaguchi; Hiroshi Kawasaki
;;;
Poster
main
https://openaccess.thecvf.com/content/WACV2025/html/Ichimaru_Neural_SDF_for_Shadow-Aware_Unsupervised_Structured_Light_WACV_2025_paper.html
0
Neural SDF for Shadow-Aware Unsupervised Structured Light Among various active 3D measurement techniques Structured Light (SL) is one of the most popular method for its robustness and high accuracy. Ordinary SL system consists of a camera and a projector and by projecting a pre-defined pattern we can obtain pixel-to-pi...
[ -0.0837111845612526, 0.002710555447265506, -0.0437379851937294, -0.02976766601204872, 0.006053189747035503, 0.029509298503398895, 0.0015132974367588758, 0.019802050665020943, -0.002265332266688347, 0.010205529630184174, 0.013638131320476532, 0.016867728903889656, -0.014044137671589851, 0.0...
wacv_2025_7a8bb80720
7a8bb80720
wacv
2,025
NeuroViG - Integrating Event Cameras for Resource-Efficient Video Grounding
Spatio-Temporal Video Grounding (STVG) - the task of identifying the target object in the field-of-view that the language instruction refers to - is a fundamental vision-language task. Current STVG approaches typically utilize feeds from an RGB camera that is assumed to be always-on and process the video frames using c...
Dulanga Weerakoon; Vigneshwaran Subbaraju; Joo Hwee Lim; Archan Misra
SMART Centre; A*STAR, Singapore; A*STAR, Singapore; Singapore Management University
Poster
main
https://openaccess.thecvf.com/content/WACV2025/html/Weerakoon_NeuroViG_-_Integrating_Event_Cameras_for_Resource-Efficient_Video_Grounding_WACV_2025_paper.html
0
NeuroViG - Integrating Event Cameras for Resource-Efficient Video Grounding Spatio-Temporal Video Grounding (STVG) - the task of identifying the target object in the field-of-view that the language instruction refers to - is a fundamental vision-language task. Current STVG approaches typically utilize feeds from an RGB...
[ -0.04630504548549652, -0.03466452285647392, -0.0025337415281683207, 0.03604155406355858, 0.010070704855024815, -0.016515221446752548, 0.036096636205911636, 0.00557698542252183, -0.00917104259133339, 0.023648254573345184, 0.014091642573475838, -0.0450565330684185, 0.0014435900375247002, 0.0...
wacv_2025_a7bf5b50fc
a7bf5b50fc
wacv
2,025
No Annotations for Object Detection in Art through Stable Diffusion
Object detection in art is a valuable tool for the digital humanities as it allows for faster identification of objects in artistic and historical images compared to humans. However annotating such images poses significant challenges due to the need for specialized domain expertise. We present NADA (no annotations for ...
Patrick Ramos; Nicolas Gonthier; Selina Khan; Yuta Nakashima; Noa Garcia
Osaka University; Univ Gustave Eiffel, ENSG, IGN, LASTIG, France; University of Amsterdam; Osaka University; Osaka University
Poster
main
https://github.com/patrick-john-ramos/nada
https://openaccess.thecvf.com/content/WACV2025/html/Ramos_No_Annotations_for_Object_Detection_in_Art_through_Stable_Diffusion_WACV_2025_paper.html
0
2412.06286
No Annotations for Object Detection in Art through Stable Diffusion Object detection in art is a valuable tool for the digital humanities as it allows for faster identification of objects in artistic and historical images compared to humans. However annotating such images poses significant challenges due to the need fo...
[ -0.08987852185964584, -0.04259491711854935, -0.03808656707406044, 0.04937548190355301, 0.021008925512433052, -0.019295750185847282, 0.0065686702728271484, 0.05878892168402672, 0.03666192665696144, 0.01733010821044445, -0.01981871947646141, -0.05997912958264351, 0.0019318292615935206, -0.00...
wacv_2025_1435ef1218
1435ef1218
wacv
2,025
Noise-Aware Evaluation of Object Detectors
Supervised object detection requires annotated datasets for training and evaluation purposes. However human annotation of large datasets is error-prone and frequent mistakes are erroneous labels missing objects and imprecise bounding boxes. The main goals of this work are to quantify the extent of annotation noise in t...
Jeffri Murrugarra Llerena; Claudio R. Jung
Computer Science Department, Stony Brook University; Institute of Informatics, Federal University of Rio Grande do Sul
Poster
main
https://github.com/Artcs1/Error-Aware
https://openaccess.thecvf.com/content/WACV2025/html/Llerena_Noise-Aware_Evaluation_of_Object_Detectors_WACV_2025_paper.html
0
Noise-Aware Evaluation of Object Detectors Supervised object detection requires annotated datasets for training and evaluation purposes. However human annotation of large datasets is error-prone and frequent mistakes are erroneous labels missing objects and imprecise bounding boxes. The main goals of this work are to q...
[ -0.0596868135035038, -0.05333952233195305, -0.048514094203710556, 0.04691799357533455, 0.0201368760317564, -0.036840274930000305, 0.010151955299079418, 0.023848742246627808, -0.009493098594248295, 0.026391370221972466, -0.004973900970071554, -0.04587867110967636, -0.0007881916826590896, -0...
wacv_2025_2458993a5a
2458993a5a
wacv
2,025
Non-Cross Diffusion for Semantic Consistency
In diffusion models deviations from a straight generative flow are a common issue resulting in semantic inconsistencies and suboptimal generations. To address this challenge we introduce `Non-Cross Diffusion' an innovative approach in generative modeling for learning ordinary differential equation (ODE) models. Our met...
Ziyang Zheng; Ruiyuan Gao; Qiang Xu
The Chinese University of Hong Kong; The Chinese University of Hong Kong; The Chinese University of Hong Kong
Poster
main
https://openaccess.thecvf.com/content/WACV2025/html/Zheng_Non-Cross_Diffusion_for_Semantic_Consistency_WACV_2025_paper.html
0
2312.00820
Non-Cross Diffusion for Semantic Consistency In diffusion models deviations from a straight generative flow are a common issue resulting in semantic inconsistencies and suboptimal generations. To address this challenge we introduce `Non-Cross Diffusion' an innovative approach in generative modeling for learning ordinar...
[ -0.03999378904700279, -0.025106800720095634, -0.027634084224700928, -0.005363557953387499, -0.013890831731259823, -0.04814750328660011, -0.007793992292135954, 0.013595674186944962, -0.001576092792674899, 0.031065285205841064, -0.0499553419649601, -0.04150646924972534, 0.008421201258897781, ...
wacv_2025_0299b8516a
0299b8516a
wacv
2,025
Now You See Me: Context-Aware Automatic Audio Description
Audio Description (AD) plays a pivotal role as an application system aimed at guaranteeing accessibility in multimedia content which provides additional narrations at suitable intervals to describe visual elements catering specifically to the needs of visually impaired audiences. In this paper we introduce CA3D the pio...
Seon-Ho Lee; Jue Wang; David Fan; Zhikang Zhang; Linda Liu; Xiang Hao; Vimal Bhat; Xinyu Li
Korea University*; Amazon AGI; Meta FAIR†; Amazon AGI; Amazon Prime Video; Amazon Prime Video; Amazon Prime Video; Amazon AGI
Poster
main
https://openaccess.thecvf.com/content/WACV2025/html/Lee_Now_You_See_Me_Context-Aware_Automatic_Audio_Description_WACV_2025_paper.html
2
Now You See Me: Context-Aware Automatic Audio Description Audio Description (AD) plays a pivotal role as an application system aimed at guaranteeing accessibility in multimedia content which provides additional narrations at suitable intervals to describe visual elements catering specifically to the needs of visually i...
[ -0.031430814415216446, -0.02262645773589611, -0.05390804633498192, 0.0012054697144776583, -0.024323906749486923, -0.011854174546897411, -0.0009198409970849752, 0.0406268946826458, -0.005992373917251825, 0.03773563355207443, 0.0006027348572388291, -0.07002449780702591, 0.006113620009273291, ...
wacv_2025_93996c1055
93996c1055
wacv
2,025
OPTIMUS: Observing Persistent Transformations in Multi-Temporal Unlabeled Satellite-Data
In the face of pressing environmental issues in the 21st century monitoring surface changes on Earth is more important than ever. Large-scale remote sensing such as satellite imagery is an important tool for this task. However using supervised methods to detect changes is difficult because of the lack of satellite data...
Raymond Yu; Paul Han; Piper Wolters; Favyen Bastani
University of Washington; University of Washington; Allen Institute for AI; Allen Institute for AI
Poster
main
https://openaccess.thecvf.com/content/WACV2025/html/Yu_OPTIMUS_Observing_Persistent_Transformations_in_Multi-Temporal_Unlabeled_Satellite-Data_WACV_2025_paper.html
0
OPTIMUS: Observing Persistent Transformations in Multi-Temporal Unlabeled Satellite-Data In the face of pressing environmental issues in the 21st century monitoring surface changes on Earth is more important than ever. Large-scale remote sensing such as satellite imagery is an important tool for this task. However usin...
[ -0.05225306376814842, -0.0037614793982356787, -0.04250657185912132, 0.0028095287270843983, -0.02998065948486328, -0.04573069512844086, 0.012850177474319935, 0.01934475265443325, 0.010302376933395863, 0.026923298835754395, -0.052104830741882324, -0.015194153413176537, 0.012711206451058388, ...
wacv_2025_3453b5064f
3453b5064f
wacv
2,025
ORFormer: Occlusion-Robust Transformer for Accurate Facial Landmark Detection
Although facial landmark detection (FLD) has gained significant progress existing FLD methods still suffer from performance drops on partially non-visible faces such as faces with occlusions or under extreme lighting conditions or poses. To address this issue we introduce ORFormer a novel transformer-based method that ...
Jui-Che Chiang; Hou-Ning Hu; Bo-Syuan Hou; Chia-Yu Tseng; Yu-Lun Liu; Min-Hung Chen; Yen-Yu Lin
National Yang Ming Chiao Tung University; MediaTek Inc.; National Yang Ming Chiao Tung University; National Yang Ming Chiao Tung University; National Yang Ming Chiao Tung University; NVIDIA; National Yang Ming Chiao Tung University
Poster
main
https://ben0919.github.io/ORFormer
https://openaccess.thecvf.com/content/WACV2025/html/Chiang_ORFormer_Occlusion-Robust_Transformer_for_Accurate_Facial_Landmark_Detection_WACV_2025_paper.html
0
2412.13174
ORFormer: Occlusion-Robust Transformer for Accurate Facial Landmark Detection Although facial landmark detection (FLD) has gained significant progress existing FLD methods still suffer from performance drops on partially non-visible faces such as faces with occlusions or under extreme lighting conditions or poses. To a...
[ -0.024618953466415405, -0.0038045793771743774, -0.008915828540921211, -0.024452734738588333, -0.003917701076716185, -0.0032920693047344685, 0.016271041706204414, 0.025228425860404968, 0.039043113589286804, 0.01057340670377016, 0.00594234699383378, -0.004411742556840181, -0.007752292323857546...
wacv_2025_34c7d49d00
34c7d49d00
wacv
2,025
ORID: Organ-Regional Information Driven Framework for Radiology Report Generation
The objective of Radiology Report Generation (RRG) is to automatically generate coherent textual analyses of diseases based on radiological images thereby alleviating the workload of radiologists. Current AI-based methods for RRG primarily focus on modifications to the encoder-decoder model architecture. To advance the...
Tiancheng Gu; Kaicheng Yang; Xiang An; Ziyong Feng; Dongnan Liu; Weidong Cai
University of Sydney; DeepGlint; DeepGlint; DeepGlint; University of Sydney; University of Sydney+DeepGlint
Poster
main
https://openaccess.thecvf.com/content/WACV2025/html/Gu_ORID_Organ-Regional_Information_Driven_Framework_for_Radiology_Report_Generation_WACV_2025_paper.html
2
2411.13025
ORID: Organ-Regional Information Driven Framework for Radiology Report Generation The objective of Radiology Report Generation (RRG) is to automatically generate coherent textual analyses of diseases based on radiological images thereby alleviating the workload of radiologists. Current AI-based methods for RRG primaril...
[ -0.006494762375950813, -0.0067202746868133545, -0.014468886889517307, 0.022533217445015907, -0.00833494495600462, -0.019664697349071503, -0.02985786460340023, 0.008127473294734955, 0.016002371907234192, 0.00518227880820632, -0.03460264950990677, -0.009480549022555351, -0.021793534979224205, ...
wacv_2025_b9f0427a0c
b9f0427a0c
wacv
2,025
OT-VP: Optimal Transport-Guided Visual Prompting for Test-Time Adaptation
Vision Transformers (ViTs) have demonstrated remarkable capabilities in learning representations but their performance is compromised when applied to unseen domains. Previous methods either engage in prompt learning during the training phase or modify model parameters at test time through entropy minimization. The form...
Yunbei Zhang; Akshay Mehra; Jihun Hamm
Tulane University; Tulane University; Tulane University
Poster
main
https://github.com/zybeich/OT-VP
https://openaccess.thecvf.com/content/WACV2025/html/Zhang_OT-VP_Optimal_Transport-Guided_Visual_Prompting_for_Test-Time_Adaptation_WACV_2025_paper.html
2
OT-VP: Optimal Transport-Guided Visual Prompting for Test-Time Adaptation Vision Transformers (ViTs) have demonstrated remarkable capabilities in learning representations but their performance is compromised when applied to unseen domains. Previous methods either engage in prompt learning during the training phase or m...
[ -0.0641908198595047, -0.017706548795104027, -0.004664923995733261, 0.03687947615981102, -0.019557854160666466, -0.01707417145371437, 0.0029281792230904102, -0.013582351617515087, 0.027677936479449272, 0.03779596462845802, -0.004692418500781059, -0.04069206491112709, 0.03297523781657219, 0....
wacv_2025_b942855b52
b942855b52
wacv
2,025
OTCXR: Rethinking Self-Supervised Alignment using Optimal Transport for Chest X-ray Analysis
Self-supervised learning (SSL) has emerged as a promising technique for analyzing medical modalities such as X-rays due to its ability to learn without annotations. However conventional SSL methods face challenges in achieving semantic alignment and capturing subtle details which limits their ability to accurately repr...
Vandan Gorade; Azad Singh; Deepak Mishra
Northwestern University; Indian Institute of Technology Jodhpur; Indian Institute of Technology Jodhpur
Poster
main
https://openaccess.thecvf.com/content/WACV2025/html/Gorade_OTCXR_Rethinking_Self-Supervised_Alignment_using_Optimal_Transport_for_Chest_X-ray_WACV_2025_paper.html
0
OTCXR: Rethinking Self-Supervised Alignment using Optimal Transport for Chest X-ray Analysis Self-supervised learning (SSL) has emerged as a promising technique for analyzing medical modalities such as X-rays due to its ability to learn without annotations. However conventional SSL methods face challenges in achieving ...
[ -0.03543870896100998, -0.015794744715094566, -0.03321800380945206, 0.01601681485772133, -0.010317012667655945, 0.0116124227643013, -0.0073653291910886765, -0.016201874241232872, 0.023743009194731712, 0.04830027371644974, 0.010585347190499306, -0.01653497852385044, -0.004126804880797863, 0....
wacv_2025_9282da860f
9282da860f
wacv
2,025
OccFlowNet: Occupancy Estimation via Differentiable Rendering and Occupancy Flow
Semantic occupancy has recently gained significant traction as a prominent 3D scene representation. However most existing camera-based methods rely on large and costly datasets with fine-grained 3D voxel labels for training which limits their practicality and scalability. Furthermore approaches in this domain lack the ...
Simon Boeder; Benjamin Risse
;
Poster
main
https://openaccess.thecvf.com/content/WACV2025/html/Boeder_OccFlowNet_Occupancy_Estimation_via_Differentiable_Rendering_and_Occupancy_Flow_WACV_2025_paper.html
0
OccFlowNet: Occupancy Estimation via Differentiable Rendering and Occupancy Flow Semantic occupancy has recently gained significant traction as a prominent 3D scene representation. However most existing camera-based methods rely on large and costly datasets with fine-grained 3D voxel labels for training which limits th...
[ -0.09707889705896378, -0.006794234737753868, -0.04000180587172508, -0.0036685189697891474, -0.020755305886268616, -0.006936835590749979, -0.011693259701132774, -0.003337317146360874, 0.0394866019487381, 0.020589705556631088, 0.015327278524637222, -0.02200651355087757, 0.007033436093479395, ...
wacv_2025_ee9a68946c
ee9a68946c
wacv
2,025
OccLoff: Learning Optimized Feature Fusion for 3D Occupancy Prediction
3D semantic occupancy prediction is crucial for finely representing the surrounding environment which is essential for ensuring the safety in autonomous driving. Existing fusion-based occupancy methods typically involve performing a 2D-to-3D view transformation on image features followed by computationally intensive 3D...
Ji Zhang; Yiran Ding; Zixin Liu
Wuhan University; Wuhan University; Wuhan University
Poster
main
https://openaccess.thecvf.com/content/WACV2025/html/Zhang_OccLoff_Learning_Optimized_Feature_Fusion_for_3D_Occupancy_Prediction_WACV_2025_paper.html
2
2411.03696
OccLoff: Learning Optimized Feature Fusion for 3D Occupancy Prediction 3D semantic occupancy prediction is crucial for finely representing the surrounding environment which is essential for ensuring the safety in autonomous driving. Existing fusion-based occupancy methods typically involve performing a 2D-to-3D view tr...
[ -0.03339720517396927, 0.01896563172340393, -0.022098923102021217, -0.0038221552968025208, -0.02585887350142002, -0.014339418150484562, -0.013040022924542427, -0.009750067256391048, 0.05230754241347313, 0.023794587701559067, -0.024660849943757057, -0.0004688418412115425, 0.02230166457593441, ...
wacv_2025_5b8bccc6ee
5b8bccc6ee
wacv
2,025
OmniDiffusion: Reformulating 360 Monocular Depth Estimation using Semantic and Surface Normal Conditioned Diffusion
Depth estimation is the fundamental computer vision task for scene analysis. With the emergence of the deep learning era supervised monocular image depth estimation (MDE) became a popular choice for the task. Predominantly MDE methods utilize 360 images as ideal input due to their comprehensive field of view scene cont...
Payal Mohadikar; Ye Duan
University of Missouri, Missouri, USA; Clemson University, South Carolina, USA
Poster
main
https://openaccess.thecvf.com/content/WACV2025/html/Mohadikar_OmniDiffusion_Reformulating_360_Monocular_Depth_Estimation_using_Semantic_and_Surface_WACV_2025_paper.html
0
OmniDiffusion: Reformulating 360 Monocular Depth Estimation using Semantic and Surface Normal Conditioned Diffusion Depth estimation is the fundamental computer vision task for scene analysis. With the emergence of the deep learning era supervised monocular image depth estimation (MDE) became a popular choice for the t...
[ -0.05625392869114876, -0.021081462502479553, -0.013256184756755829, -0.015586338005959988, -0.00008872814214555547, 0.006774873472750187, -0.016035854816436768, 0.013852483592927456, 0.05603375658392906, 0.02209058403968811, -0.03640175983309746, -0.038933735340833664, -0.011137030087411404,...
wacv_2025_a0d6df9fb9
a0d6df9fb9
wacv
2,025
OmniGS: Fast Radiance Field Reconstruction using Omnidirectional Gaussian Splatting
Photorealistic reconstruction relying on 3D Gaussian Splatting has shown promising potential in various domains. However the current 3D Gaussian Splatting system only supports radiance field reconstruction using undistorted perspective images. In this paper we present OmniGS a novel omnidirectional Gaussian splatting s...
Longwei Li; Huajian Huang; Sai-Kit Yeung; Hui Cheng
Sun Yat-sen University; The Hong Kong University of Science and Technology; The Hong Kong University of Science and Technology; Sun Yat-sen University
Poster
main
https://github.com/liquorleaf/OmniGS
https://openaccess.thecvf.com/content/WACV2025/html/Li_OmniGS_Fast_Radiance_Field_Reconstruction_using_Omnidirectional_Gaussian_Splatting_WACV_2025_paper.html
2
2404.03202
OmniGS: Fast Radiance Field Reconstruction using Omnidirectional Gaussian Splatting Photorealistic reconstruction relying on 3D Gaussian Splatting has shown promising potential in various domains. However the current 3D Gaussian Splatting system only supports radiance field reconstruction using undistorted perspective ...
[ -0.10527830570936203, 0.02823239006102085, -0.01997418701648712, -0.0368400514125824, -0.010952696204185486, -0.030678585171699524, -0.018760286271572113, 0.0052464427426457405, 0.03393404558300972, 0.04344293475151062, -0.04840889200568199, 0.013546030037105083, 0.012883901596069336, 0.02...
wacv_2025_b89dcee1e1
b89dcee1e1
wacv
2,025
On Explaining Knowledge Distillation: Measuring and Visualising the Knowledge Transfer Process
Knowledge distillation (KD) remains challenging due to the opaque nature of the knowledge transfer process from a Teacher to a Student making it difficult to address certain issues related to KD. To address this we proposed UniCAM a novel gradient-based visual explanation method which effectively interprets the knowled...
Gereziher Adhane; Mohammad Mahdi Dehshibi; Dennis Vetter; David Masip; Gemma Roig
Universitat Oberta de Catalunya; Universidad Carlos III de Madrid; Goethe University Frankfurt; Universitat Oberta de Catalunya; Goethe University Frankfurt
Poster
main
https://openaccess.thecvf.com/content/WACV2025/html/Adhane_On_Explaining_Knowledge_Distillation_Measuring_and_Visualising_the_Knowledge_Transfer_WACV_2025_paper.html
0
2412.13943
On Explaining Knowledge Distillation: Measuring and Visualising the Knowledge Transfer Process Knowledge distillation (KD) remains challenging due to the opaque nature of the knowledge transfer process from a Teacher to a Student making it difficult to address certain issues related to KD. To address this we proposed U...
[ -0.05751959979534149, -0.021447986364364624, -0.04443332925438881, -0.029116017743945122, -0.023322811350226402, -0.03515294939279556, 0.011014590971171856, 0.008853856474161148, 0.007850825786590576, 0.044808294624090195, -0.0015127486549317837, -0.0011711791157722473, -0.027147451415657997...
wacv_2025_b8e15f58f7
b8e15f58f7
wacv
2,025
On Neural BRDFs: A Thorough Comparison of State-of-the-Art Approaches
The bidirectional reflectance distribution function (BRDF) is an essential tool to capture the complex interaction of light and matter. Recently several works have employed neural methods for BRDF modeling following various strategies ranging from utilizing existing parametric models to purely neural parametrizations. ...
Florian Hofherr; Bjoern Haefner; Daniel Cremers
Technical University of Munich + Munich Center for Machine Learning + NVIDIA; Technical University of Munich + Munich Center for Machine Learning + NVIDIA; Technical University of Munich + Munich Center for Machine Learning
Poster
main
https://openaccess.thecvf.com/content/WACV2025/html/Hofherr_On_Neural_BRDFs_A_Thorough_Comparison_of_State-of-the-Art_Approaches_WACV_2025_paper.html
0
On Neural BRDFs: A Thorough Comparison of State-of-the-Art Approaches The bidirectional reflectance distribution function (BRDF) is an essential tool to capture the complex interaction of light and matter. Recently several works have employed neural methods for BRDF modeling following various strategies ranging from ut...
[ -0.055034879595041275, -0.013103542849421501, -0.010295640677213669, -0.018382398411631584, -0.024391310289502144, -0.011044414713978767, 0.07229411602020264, -0.011980382725596428, 0.04159438982605934, 0.00984637625515461, -0.04451460763812065, -0.013038025237619877, -0.02154596894979477, ...
wacv_2025_2a95a3f18d
2a95a3f18d
wacv
2,025
On Which Data Distribution (Synthetic or Real) We Should Rely for Soft Biometric Classification
Identification of gender is critical not only for human-computer interaction but also for scrutinizing the search space in which an identity needs to be determined. Traditionally "real" facial images are employed for gender identification by computer vision algorithms. Due to the tremendous rise of privacy and advancem...
Manju R. A; Atul Kumar; Akshay Agarwal
Trustworthy BiometraVision Lab, IISER Bhopal; Trustworthy BiometraVision Lab, IISER Bhopal; Trustworthy BiometraVision Lab, IISER Bhopal
Poster
main
https://openaccess.thecvf.com/content/WACV2025/html/A_On_Which_Data_Distribution_Synthetic_or_Real_We_Should_Rely_WACV_2025_paper.html
0
On Which Data Distribution (Synthetic or Real) We Should Rely for Soft Biometric Classification Identification of gender is critical not only for human-computer interaction but also for scrutinizing the search space in which an identity needs to be determined. Traditionally "real" facial images are employed for gender ...
[ -0.06404998898506165, -0.023990655317902565, -0.030489293858408928, -0.024309031665325165, -0.023578636348247528, -0.039628591388463974, 0.01817558892071247, 0.01704254187643528, 0.05873122066259384, 0.022792058065533638, -0.03659464418888092, -0.02417793497443199, -0.039628591388463974, 0...
wacv_2025_c10d910d48
c10d910d48
wacv
2,025
On the Importance of Dual-Space Augmentation for Domain Generalized Object Detection
The distribution gap between training data and real-world data often causes significant performance drops in networks trained via naive supervised learning. To address this domain generalization methods have been developed to gain robust performance in unseen domains. In this paper we propose a single-domain generalize...
Hayoung Park; Choongsang Cho; Guisik Kim
Korea Electronics Technology Institute (KETI); Korea Electronics Technology Institute (KETI); Korea Electronics Technology Institute (KETI)
Poster
main
https://openaccess.thecvf.com/content/WACV2025/html/Park_On_the_Importance_of_Dual-Space_Augmentation_for_Domain_Generalized_Object_WACV_2025_paper.html
0
On the Importance of Dual-Space Augmentation for Domain Generalized Object Detection The distribution gap between training data and real-world data often causes significant performance drops in networks trained via naive supervised learning. To address this domain generalization methods have been developed to gain robu...
[ -0.09166911989450455, -0.03443942591547966, -0.03915715590119362, -0.024223728105425835, -0.012783233076334, -0.010034248232841492, -0.026165254414081573, 0.008242418058216572, 0.012456621043384075, 0.018145114183425903, -0.04169747233390808, -0.01790015585720539, -0.04579826816916466, 0.0...
wacv_2025_481ab6650d
481ab6650d
wacv
2,025
On-the-Fly Object-aware Representative Point Selection in Point Cloud
Point clouds are essential for object modeling and play a critical role in assisting driving tasks for autonomous vehicles (AVs). However the significant volume of data generated by AVs creates challenges for storage bandwidth and processing cost. To tackle these challenges we propose a representative point selection f...
Xiaoyu Zhang; Ziwei Wang; Hai Dong; Zhifeng Bao; Jiajun Liu
RMIT University; CSIRO; RMIT University; RMIT University; CSIRO
Poster
main
https://openaccess.thecvf.com/content/WACV2025/html/Zhang_On-the-Fly_Object-aware_Representative_Point_Selection_in_Point_Cloud_WACV_2025_paper.html
0
On-the-Fly Object-aware Representative Point Selection in Point Cloud Point clouds are essential for object modeling and play a critical role in assisting driving tasks for autonomous vehicles (AVs). However the significant volume of data generated by AVs creates challenges for storage bandwidth and processing cost. To...
[ -0.039393290877342224, -0.02404477261006832, -0.026478979736566544, 0.018591031432151794, -0.013555379584431648, -0.02922907844185829, 0.05106262490153313, 0.001017350354231894, 0.00116252014413476, 0.051657240837812424, -0.012189622037112713, -0.036383047699928284, 0.01943650096654892, 0....
wacv_2025_774c460e49
774c460e49
wacv
2,025
One VLM to Keep it Learning: Generation and Balancing for Data-Free Continual Visual Question Answering
Vision-Language Models (VLMs) have shown significant promise in Visual Question Answering (VQA) tasks by leveraging web-scale multimodal datasets. However these models often struggle with continual learning due to catastrophic forgetting when adapting to new tasks. As an effective remedy to mitigate catastrophic forget...
Deepayan Das; Davide Talon; Massimiliano Mancini; Yiming Wang; Elisa Ricci
University of Trento; Fondazione Bruno Kessler; University of Trento; Fondazione Bruno Kessler; University of Trento+Fondazione Bruno Kessler
Poster
main
https://github.com/Deepayan137/GaB
https://openaccess.thecvf.com/content/WACV2025/html/Das_One_VLM_to_Keep_it_Learning_Generation_and_Balancing_for_WACV_2025_paper.html
1
2411.02210
One VLM to Keep it Learning: Generation and Balancing for Data-Free Continual Visual Question Answering Vision-Language Models (VLMs) have shown significant promise in Visual Question Answering (VQA) tasks by leveraging web-scale multimodal datasets. However these models often struggle with continual learning due to ca...
[ -0.028125813230872154, -0.022593475878238678, -0.021238237619400024, 0.010154996998608112, -0.009523790329694748, 0.006483789067715406, 0.05380105972290039, 0.04949400573968887, -0.029833784326910973, 0.0005453437333926558, -0.025600988417863846, -0.019196100533008575, -0.018304985016584396,...
wacv_2025_c7f2f503aa
c7f2f503aa
wacv
2,025
Online-LoRA: Task-Free Online Continual Learning via Low Rank Adaptation
Catastrophic forgetting is a significant challenge in online continual learning (OCL) especially for non-stationary data streams that do not have well-defined task boundaries. This challenge is exacerbated by the memory constraints and privacy concerns inherent in rehearsal buffers. To tackle catastrophic forgetting in...
Xiwen Wei; Guihong Li; Radu Marculescu
The University of Texas at Austin; The University of Texas at Austin; The University of Texas at Austin
Poster
main
https://github.com/Christina200/Online-LoRA-official.git
https://openaccess.thecvf.com/content/WACV2025/html/Wei_Online-LoRA_Task-Free_Online_Continual_Learning_via_Low_Rank_Adaptation_WACV_2025_paper.html
4
Online-LoRA: Task-Free Online Continual Learning via Low Rank Adaptation Catastrophic forgetting is a significant challenge in online continual learning (OCL) especially for non-stationary data streams that do not have well-defined task boundaries. This challenge is exacerbated by the memory constraints and privacy con...
[ -0.07834495604038239, -0.0007463853107765317, -0.013746637850999832, 0.017110755667090416, -0.019114717841148376, 0.012994018383324146, 0.06873319298028946, 0.029796471819281578, 0.02801920287311077, -0.01439044438302517, 0.0007248495239764452, -0.04417784884572029, -0.00038027678965590894, ...
wacv_2025_4ddcc4a0e8
4ddcc4a0e8
wacv
2,025
OpenCapBench: A Benchmark to Bridge Pose Estimation and Biomechanics
Pose estimation has promised to impact healthcare by enabling more practical methods to quantify nuances of human movement and biomechanics. However despite the inherent connection between pose estimation and biomechanics these disciplines have largely remained disparate. For example most current pose estimation benchm...
Yoni Gozlan; Antoine Falisse; Scott Uhlrich; Anthony Gatti; Michael Black; Jennifer Hicks; Scott Delp; Akshay Chaudhari
Stanford University; Stanford University; Stanford University; Stanford University; Max Planck Institute for Intelligent Systems; Stanford University; Stanford University; Stanford University
Poster
main
https://openaccess.thecvf.com/content/WACV2025/html/Gozlan_OpenCapBench_A_Benchmark_to_Bridge_Pose_Estimation_and_Biomechanics_WACV_2025_paper.html
6
2406.09788
OpenCapBench: A Benchmark to Bridge Pose Estimation and Biomechanics Pose estimation has promised to impact healthcare by enabling more practical methods to quantify nuances of human movement and biomechanics. However despite the inherent connection between pose estimation and biomechanics these disciplines have largel...
[ 0.018364278599619865, -0.021977534517645836, -0.05937659367918968, -0.010821142233908176, 0.002222198760136962, -0.0047563668340444565, 0.04101231321692467, 0.004698163364082575, 0.02115803211927414, -0.002033620374277234, -0.03220267593860626, 0.009619827382266521, -0.0018671591533347964, ...
wacv_2025_7dbb8b6ad8
7dbb8b6ad8
wacv
2,025
OpenCity3D: What do Vision-Language Models Know About Urban Environments?
The rise of 2D vision-language models (VLMs) has enabled new possibilities for language-driven 3D scene understanding tasks. Existing works focus on indoor scenes or autonomous driving scenarios and typically validate against a pre-defined set of semantic object classes. In this work we analyze the capabilities of visi...
Valentin Bieri; Marco Zamboni; Nicolas Samuel Blumer; Qingxuan Chen; Francis Engelmann
ETH Zürich; ETH Zürich; ETH Zürich+University of Zurich; ETH Zürich+University of Zurich; ETH Zürich+Stanford University
Poster
main
https://github.com/opencity3d
https://openaccess.thecvf.com/content/WACV2025/html/Bieri_OpenCity3D_What_do_Vision-Language_Models_Know_About_Urban_Environments_WACV_2025_paper.html
1
OpenCity3D: What do Vision-Language Models Know About Urban Environments? The rise of 2D vision-language models (VLMs) has enabled new possibilities for language-driven 3D scene understanding tasks. Existing works focus on indoor scenes or autonomous driving scenarios and typically validate against a pre-defined set of...
[ -0.04313633218407631, 0.006766483653336763, -0.007957053370773792, -0.004353165626525879, -0.007446809206157923, 0.021347520872950554, 0.002838521497324109, 0.023811401799321175, 0.019324930384755135, -0.02362753078341484, -0.04126083850860596, 0.00621946481987834, 0.010811664164066315, -0...
wacv_2025_da948ed3dd
da948ed3dd
wacv
2,025
Optimizing Dense Visual Predictions Through Multi-Task Coherence and Prioritization
Multi-Task Learning (MTL) involves the concurrent training of multiple tasks offering notable advantages for dense prediction tasks in computer vision. MTL not only reduces training and inference time as opposed to having multiple single-task models but also enhances task accuracy through the interaction of multiple ta...
Maxime Fontana; Michael Spratling; Miaojing Shi
Department of Informatics, King’s College London; Department of Behavioural and Cognitive Sciences, University of Luxembourg; College of Electronic and Information Engineering, Tongji University
Poster
main
https://github.com/Klodivio355/MT-CP
https://openaccess.thecvf.com/content/WACV2025/html/Fontana_Optimizing_Dense_Visual_Predictions_Through_Multi-Task_Coherence_and_Prioritization_WACV_2025_paper.html
0
2412.03179
Optimizing Dense Visual Predictions Through Multi-Task Coherence and Prioritization Multi-Task Learning (MTL) involves the concurrent training of multiple tasks offering notable advantages for dense prediction tasks in computer vision. MTL not only reduces training and inference time as opposed to having multiple singl...
[ -0.033700644969940186, -0.0369572751224041, -0.019430013373494148, 0.03419462963938713, -0.0002521373680792749, 0.030297648161649704, 0.014041935093700886, 0.0014007630525156856, 0.004944407846778631, 0.029730482026934624, -0.043982818722724915, -0.02852296642959118, -0.005232564639300108, ...
wacv_2025_3cecadcd51
3cecadcd51
wacv
2,025
Optimizing Neural Network Effectiveness via Non-Monotonicity Refinement
Activation functions play a crucial role in artificial neural networks by introducing non-linearities that enable networks to learn complex patterns in data. An appropriate choice of an activation function plays a crucial role in the training dynamics of a neural network which can boost network performance significantl...
Koushik Biswas; Amit Reza; Meghana Karri; Debesh Jha; Hongyi Pan; Nikhil Tomar; Aliza Subedi; Smriti Regmi; Ulas Bagci
;;;;;;;;
Poster
main
https://github.com/koushik313/AMSU
https://openaccess.thecvf.com/content/WACV2025/html/Biswas_Optimizing_Neural_Network_Effectiveness_via_Non-Monotonicity_Refinement_WACV_2025_paper.html
0
Optimizing Neural Network Effectiveness via Non-Monotonicity Refinement Activation functions play a crucial role in artificial neural networks by introducing non-linearities that enable networks to learn complex patterns in data. An appropriate choice of an activation function plays a crucial role in the training dynam...
[ 0.023768434301018715, -0.028462747111916542, -0.04608961194753647, -0.027386581525206566, -0.012515073642134666, -0.003613507142290473, 0.035216618329286575, 0.025753777474164963, 0.008261432871222496, 0.039595503360033035, -0.014982834458351135, -0.0429353304207325, -0.04367751628160477, ...
wacv_2025_5555b2e676
5555b2e676
wacv
2,025
Optimizing Vision-Language Model for Road Crossing Intention Estimation
Identifying a pedestrian's intention to cross the road is crucial for autonomous driving as it alerts the system to stop or slow down. However determining crossing intention from video is challenging due to the need for extracting complex high-level semantics. This paper introduces ClipCross a novel classification fram...
Roy Uziel; Oded Bialer
General Motors, Technical Center Israel + Ben-Gurion University of the Negev; General Motors, Technical Center Israel
Poster
main
https://openaccess.thecvf.com/content/WACV2025/html/Uziel_Optimizing_Vision-Language_Model_for_Road_Crossing_Intention_Estimation_WACV_2025_paper.html
0
Optimizing Vision-Language Model for Road Crossing Intention Estimation Identifying a pedestrian's intention to cross the road is crucial for autonomous driving as it alerts the system to stop or slow down. However determining crossing intention from video is challenging due to the need for extracting complex high-leve...
[ -0.02836393006145954, -0.055780574679374695, -0.00045713380677625537, 0.03594221919775009, -0.01960153318941593, -0.01855405420064926, 0.002170104766264558, -0.02000230923295021, -0.009691464714705944, 0.03208020702004433, -0.013517042621970177, -0.03180694952607155, 0.02433796413242817, 0...
wacv_2025_a04bba180b
a04bba180b
wacv
2,025
Ordinal Multiple-Instance Learning for Ulcerative Colitis Severity Estimation with Selective Aggregated Transformer
Patient-level diagnosis of severity in ulcerative colitis (UC) is common in clinical practice where the most severe score for a patient is typically recorded as the diagnosis result. However previous UC classification methods (i.e. image-level estimation) mainly assumed the input was a single image. Thus these methods ...
Kaito Shiku; Kazuya Nishimura; Daiki Suehiro; Kiyohito Tanaka; Ryoma Bise
Kyushu University; National Cancer Center Japan; Kyushu University; Kyoto Second Red Cross Hospital; Kyushu University
Poster
main
https://github.com/Shiku-Kaito/Ordinal-Multiple-instance-Learning-for-Ulcerative-Colitis-Severity-Estimation
https://openaccess.thecvf.com/content/WACV2025/html/Shiku_Ordinal_Multiple-Instance_Learning_for_Ulcerative_Colitis_Severity_Estimation_with_Selective_WACV_2025_paper.html
0
2411.14750
Ordinal Multiple-Instance Learning for Ulcerative Colitis Severity Estimation with Selective Aggregated Transformer Patient-level diagnosis of severity in ulcerative colitis (UC) is common in clinical practice where the most severe score for a patient is typically recorded as the diagnosis result. However previous UC c...
[ -0.05015316233038902, -0.008829781785607338, -0.05677084997296333, 0.025541305541992188, 0.010196074843406677, -0.010651505552232265, -0.011878380551934242, 0.009787116199731827, 0.02122865431010723, 0.013244672678411007, 0.005358283407986164, 0.014620260335505009, 0.019760122522711754, 0....
wacv_2025_e7e3294c1b
e7e3294c1b
wacv
2,025
Oriented Cell Dataset: A Dataset and Benchmark for Oriented Cell Detection and Applications
This work presents a new public dataset for cell detection in bright-field microscopy images annotated with Oriented Bounding Boxes (OBBs) named Oriented Cell Dataset (OCD). Our dataset also contains a subset of images with five independent expert annotations which allows inter-annotation analysis to determine a suitab...
Lucas Kirsten; Angelo Angonezi; Jose Marques; Fernanda Oliveira; Juliano Faccioni; Camila Cassel; Débora de Sousa; Samlai Vedovatto; Guido Lenz; Claudio Jung
Federal University of Rio Grande do Sul, Brazil – Institute of Informatics; Federal University of Rio Grande do Sul, Brazil – Institute of Biosciences; Federal University of Rio Grande do Sul, Brazil – Institute of Informatics; Federal University of Rio Grande do Sul, Brazil – Institute of Biosciences; Federal Universi...
Poster
main
https://github.com/LucasKirsten/Deep-Cell-Tracking-EBB
https://openaccess.thecvf.com/content/WACV2025/html/Kirsten_Oriented_Cell_Dataset_A_Dataset_and_Benchmark_for_Oriented_Cell_WACV_2025_paper.html
0
Oriented Cell Dataset: A Dataset and Benchmark for Oriented Cell Detection and Applications This work presents a new public dataset for cell detection in bright-field microscopy images annotated with Oriented Bounding Boxes (OBBs) named Oriented Cell Dataset (OCD). Our dataset also contains a subset of images with five...
[ -0.03516290336847305, -0.031905706971883774, -0.08601956814527512, 0.0008929526666179299, -0.010169481858611107, -0.03719865158200264, 0.01310280803591013, -0.0030281736981123686, 0.022337697446346283, 0.030813807621598244, -0.03538498282432556, -0.03246091306209564, 0.02255977876484394, 0...
wacv_2025_19839217f5
19839217f5
wacv
2,025
PACA: Prespective-Aware Cross-Attention Representation for Zero-Shot Scene Rearrangement
Scene rearrangement like table tidying is a challenging task in robotic manipulation due to the complexity of predicting diverse object arrangements. Web-scale trained generative models such as Stable Diffusion can aid by generating natural scenes as goals. To facilitate robot execution object-level representations mus...
Shutong Jin; Ruiyu Wang; Kuangyi Chen; Florian T. Pokorny
KTH Royal Institute of Technology; KTH Royal Institute of Technology; Graz University of Technology; KTH Royal Institute of Technology
Poster
main
https://openaccess.thecvf.com/content/WACV2025/html/Jin_PACA_Prespective-Aware_Cross-Attention_Representation_for_Zero-Shot_Scene_Rearrangement_WACV_2025_paper.html
0
PACA: Prespective-Aware Cross-Attention Representation for Zero-Shot Scene Rearrangement Scene rearrangement like table tidying is a challenging task in robotic manipulation due to the complexity of predicting diverse object arrangements. Web-scale trained generative models such as Stable Diffusion can aid by generatin...
[ -0.05964406207203865, -0.021687103435397148, -0.04257701709866524, 0.033047012984752655, 0.006531495600938797, 0.002366648055613041, 0.03417032212018967, 0.05442611128091812, 0.03440585359930992, 0.06178197264671326, -0.025763625279068947, -0.009819891303777695, 0.0054670702666044235, 0.01...
wacv_2025_9e834e008a
9e834e008a
wacv
2,025
PALO: A Polyglot Large Multimodal Model for 5B People
In pursuit of more inclusive Vision-Language Models (VLMs) this study introduces a Large Multilingual Multimodal Model called PALO. PALO offers visual reasoning capabilities in 10 major languages including English Chinese Hindi Spanish French Arabic Bengali Russian Urdu and Japanese that span a total of 5B people (65%...
Hanoona Rasheed; Muhammad Maaz; Abdelrahman Shaker; Salman Khan; Hisham Cholakkal; Rao M. Anwer; Tim Baldwin; Michael Felsberg; Fahad S. Khan
Mohamed bin Zayed University of AI; Mohamed bin Zayed University of AI; Mohamed bin Zayed University of AI; Mohamed bin Zayed University of AI+Australian National University; Mohamed bin Zayed University of AI; Mohamed bin Zayed University of AI+Aalto University; Mohamed bin Zayed University of AI+The University of Mel...
Poster
main
https://github.com/mbzuai-oryx/PALO
https://openaccess.thecvf.com/content/WACV2025/html/Rasheed_PALO_A_Polyglot_Large_Multimodal_Model_for_5B_People_WACV_2025_paper.html
16
2402.14818
PALO: A Polyglot Large Multimodal Model for 5B People In pursuit of more inclusive Vision-Language Models (VLMs) this study introduces a Large Multilingual Multimodal Model called PALO. PALO offers visual reasoning capabilities in 10 major languages including English Chinese Hindi Spanish French Arabic Bengali Russian ...
[ 0.00557813560590148, -0.0399758443236351, -0.005163705442100763, 0.025723256170749664, -0.025418387725949287, 0.01015592273324728, 0.021131178364157677, -0.013842922635376453, -0.02101685293018818, -0.021645642817020416, -0.003999013453722, 0.009441387839615345, 0.012537704780697823, 0.035...
wacv_2025_6e2b5ebf13
6e2b5ebf13
wacv
2,025
PC-GZSL: Prior Correction for Generalized Zero Shot Learning
Generalized Zero Shot Learning (GZSL) aims at achieving a good accuracy on both seen and unseen classes by relying on the information acquired from auxiliary attributes. Existing approaches have devised many frameworks to make this knowledge transfer more efficient and informative. Despite their effectiveness on boosti...
S Divakar Bhat; Amit More; Mudit Soni; Bhuvan Aggarwal
;;;
Poster
main
https://openaccess.thecvf.com/content/WACV2025/html/Bhat_PC-GZSL_Prior_Correction_for_Generalized_Zero_Shot_Learning_WACV_2025_paper.html
0
PC-GZSL: Prior Correction for Generalized Zero Shot Learning Generalized Zero Shot Learning (GZSL) aims at achieving a good accuracy on both seen and unseen classes by relying on the information acquired from auxiliary attributes. Existing approaches have devised many frameworks to make this knowledge transfer more eff...
[ -0.04652048274874687, -0.02427811361849308, -0.03392903879284859, -0.0032562455162405968, 0.02292095124721527, -0.02082866057753563, 0.031214717775583267, 0.0188023429363966, -0.014146524481475353, 0.014909928664565086, -0.026332704350352287, -0.007982749491930008, -0.014372718520462513, 0...
wacv_2025_89537271ed
89537271ed
wacv
2,025
PETALface: Parameter Efficient Transfer Learning for Low-Resolution Face Recognition
Pre-training on large-scale datasets and utilizing margin-based loss functions have been highly successful in training models for high-resolution face recognition. However these models struggle with low-resolution face datasets in which the faces lack the facial attributes necessary for distinguishing different faces. ...
Kartik Narayan; Nithin Gopalakrishnan Nair; Jennifer Xu; Rama Chellappa; Vishal M. Patel
Johns Hopkins University; Johns Hopkins University; Systems and Technology Research; Johns Hopkins University; Johns Hopkins University
Poster
main
https://kartik-3004.github.io/PETALface/
https://openaccess.thecvf.com/content/WACV2025/html/Narayan_PETALface_Parameter_Efficient_Transfer_Learning_for_Low-Resolution_Face_Recognition_WACV_2025_paper.html
2
2412.07771
PETALface: Parameter Efficient Transfer Learning for Low-Resolution Face Recognition Pre-training on large-scale datasets and utilizing margin-based loss functions have been highly successful in training models for high-resolution face recognition. However these models struggle with low-resolution face datasets in whic...
[ -0.038645680993795395, 0.009706906974315643, -0.03369670361280441, -0.016702795401215553, -0.014364766888320446, -0.010916858911514282, 0.06768452376127243, 0.01963215135037899, 0.006236255634576082, 0.005176410544663668, -0.05582154169678688, 0.0038231750950217247, 0.010753106325864792, 0...
wacv_2025_c0294366df
c0294366df
wacv
2,025
PGRID: Power Grid Reconstruction in Informal Developments using High-Resolution Aerial Imagery
As of 2023 a record 117 million people have been displaced worldwide more than double the number from a decade ago [22]. Of these 32 million are refugees under the UNHCR mandate with 8.7 million residing in refugee camps. A critical issue faced by these populations is the lack of access to electricity with 80% of the 8...
Simone Fobi Nsutezo; Amrita Gupta; Duncan Kebut; Seema Iyer; Luana Marotti; Rahul Dodhia; Juan M. Lavista Ferres; Anthony Ortiz
Microsoft AI for Good Research Lab; Microsoft AI for Good Research Lab; HOTOSM; USA for UNHCR; Microsoft AI for Good Research Lab; Microsoft AI for Good Research Lab; Microsoft AI for Good Research Lab; Microsoft AI for Good Research Lab
Poster
main
https://openaccess.thecvf.com/content/WACV2025/html/Nsutezo_PGRID_Power_Grid_Reconstruction_in_Informal_Developments_using_High-Resolution_Aerial_WACV_2025_paper.html
0
2412.07944
PGRID: Power Grid Reconstruction in Informal Developments using High-Resolution Aerial Imagery As of 2023 a record 117 million people have been displaced worldwide more than double the number from a decade ago [22]. Of these 32 million are refugees under the UNHCR mandate with 8.7 million residing in refugee camps. A c...
[ -0.035104986280202866, 0.019327137619256973, -0.027405163273215294, -0.06260712444782257, -0.01986050046980381, -0.03273879364132881, 0.016049379482865334, -0.036559611558914185, -0.003961432725191116, 0.06218043342232704, -0.04193203151226044, -0.00595427118241787, -0.03403826057910919, 0...
wacv_2025_0e87746b04
0e87746b04
wacv
2,025
PICASSO: A Feed-Forward Framework for Parametric Inference of CAD Sketches via Rendering Self-Supervision
This work introduces PICASSO a framework for the parameterization of 2D CAD sketches from hand-drawn and precise sketch images. PICASSO converts a given CAD sketch image into parametric primitives that can be seamlessly integrated into CAD software. Our framework leverages rendering self-supervision to enable the pre-t...
Ahmet Serdar Karadeniz; Dimitrios Mallis; Nesryne Mejri; Kseniya Cherenkova; Anis Kacem; Djamila Aouada
;;;;;
Poster
main
https://openaccess.thecvf.com/content/WACV2025/html/Karadeniz_PICASSO_A_Feed-Forward_Framework_for_Parametric_Inference_of_CAD_Sketches_WACV_2025_paper.html
4
2407.13394
PICASSO: A Feed-Forward Framework for Parametric Inference of CAD Sketches via Rendering Self-Supervision This work introduces PICASSO a framework for the parameterization of 2D CAD sketches from hand-drawn and precise sketch images. PICASSO converts a given CAD sketch image into parametric primitives that can be seaml...
[ -0.02878701128065586, 0.003833593800663948, -0.031139185652136803, -0.020319178700447083, -0.00498932646587491, -0.019903024658560753, 0.07038433104753494, 0.02799089066684246, 0.004048455972224474, 0.053810540586709976, 0.001035296474583447, 0.004430684726685286, 0.023901723325252533, 0.0...
wacv_2025_2dbdc6c74c
2dbdc6c74c
wacv
2,025
PK-YOLO: Pretrained Knowledge Guided YOLO for Brain Tumor Detection in Multiplanar MRI Slices
Brain tumor detection in multiplane Magnetic Resonance Imaging (MRI) slices is a challenging task due to the various appearances and relationships in the structure of the multiplane images. In this paper we propose a new You Only Look Once (YOLO)-based detection model that incorporates Pretrained Knowledge (PK) called ...
Ming Kang; Fung Fung Ting; Raphael C.-W. Phan; Chee-Ming Ting
Monash University Malaysia Campus; Monash University Malaysia Campus; Monash University Malaysia Campus; Monash University Malaysia Campus
Poster
main
https://github.com/mkang315/PK-YOLO
https://openaccess.thecvf.com/content/WACV2025/html/Kang_PK-YOLO_Pretrained_Knowledge_Guided_YOLO_for_Brain_Tumor_Detection_in_WACV_2025_paper.html
0
PK-YOLO: Pretrained Knowledge Guided YOLO for Brain Tumor Detection in Multiplanar MRI Slices Brain tumor detection in multiplane Magnetic Resonance Imaging (MRI) slices is a challenging task due to the various appearances and relationships in the structure of the multiplane images. In this paper we propose a new You O...
[ -0.019256234169006348, -0.025271479040384293, -0.03313040733337402, 0.0049211410805583, 0.006927774753421545, -0.008710931986570358, 0.01443286333233118, 0.02826979011297226, -0.011685963720083237, 0.018148161470890045, -0.023837503045797348, -0.032180629670619965, -0.003068147459998727, 0...
wacv_2025_b53a7bc415
b53a7bc415
wacv
2,025
PLReMix: Combating Noisy Labels with Pseudo-Label Relaxed Contrastive Representation Learning
Recently the usage of Contrastive Representation Learning (CRL) as a pre-training technique improves the performance of learning with noisy labels (LNL) methods. However instead of pre-training when trivially combining CRL loss with LNL methods as an end-to-end framework the empirical experiments show severe degenerati...
Xiaoyu Liu; Beitong Zhou; Zuogong Yue; Cheng Cheng
Huazhong University of Science and Technology; Huazhong University of Science and Technology; Huazhong University of Science and Technology; Huazhong University of Science and Technology
Poster
main
https://github.com/lxysl/PLReMix
https://openaccess.thecvf.com/content/WACV2025/html/Liu_PLReMix_Combating_Noisy_Labels_with_Pseudo-Label_Relaxed_Contrastive_Representation_Learning_WACV_2025_paper.html
3
2402.17589
PLReMix: Combating Noisy Labels with Pseudo-Label Relaxed Contrastive Representation Learning Recently the usage of Contrastive Representation Learning (CRL) as a pre-training technique improves the performance of learning with noisy labels (LNL) methods. However instead of pre-training when trivially combining CRL los...
[ -0.08479823172092438, -0.00408514030277729, -0.02118118852376938, -0.021860897541046143, -0.01157341618090868, -0.030899183824658394, 0.02265082858502865, 0.011270302347838879, -0.009635328315198421, 0.025645220652222633, -0.042178671807050705, -0.013731949962675571, -0.03071547858417034, ...
wacv_2025_4960e46890
4960e46890
wacv
2,025
PRoGS: Progressive Rendering of Gaussian Splats
Over the past year 3D Gaussian Splatting (3DGS) has received significant attention for its ability to represent 3D scenes in a perceptually accurate manner. However it can require a substantial amount of storage since each splat's individual data must be stored. While compression techniques offer a potential solution b...
Brent Zoomers; Maarten Wijnants; Ivan Molenaers; Joni Vanherck; Jeroen Put; Lode Jorissen; Nick Michiels
Hasselt University - Flanders Make - Digital Future Lab, Diepenbeek, Belgium; Hasselt University - Flanders Make - Digital Future Lab, Diepenbeek, Belgium; Hasselt University - Flanders Make - Digital Future Lab, Diepenbeek, Belgium; Hasselt University - Flanders Make - Digital Future Lab, Diepenbeek, Belgium; Hasselt ...
Poster
main
https://openaccess.thecvf.com/content/WACV2025/html/Zoomers_PRoGS_Progressive_Rendering_of_Gaussian_Splats_WACV_2025_paper.html
2
2409.01761
PRoGS: Progressive Rendering of Gaussian Splats Over the past year 3D Gaussian Splatting (3DGS) has received significant attention for its ability to represent 3D scenes in a perceptually accurate manner. However it can require a substantial amount of storage since each splat's individual data must be stored. While com...
[ -0.059192705899477005, -0.021679094061255455, -0.028121203184127808, -0.019018223509192467, -0.025880469009280205, -0.020315982401371002, 0.003837255761027336, -0.023303626105189323, 0.0020855157636106014, 0.029783079400658607, -0.0014588107587769628, 0.02776641957461834, 0.02998848073184490...
wacv_2025_c561bb5141
c561bb5141
wacv
2,025
PTQ4VM: Post-Training Quantization for Visual Mamba
Visual Mamba is an approach that extends the selective space state model Mamba to vision tasks. It processes image tokens sequentially in a fixed order accumulating information to generate outputs. Despite its growing popularity for delivering high-quality outputs at a low computational cost across various tasks Visual...
Younghyun Cho; Changhun Lee; Seonggon Kim; Eunhyeok Park
Pohang University of Science and Technology (POSTECH), Republic of Korea; Pohang University of Science and Technology (POSTECH), Republic of Korea; Pohang University of Science and Technology (POSTECH), Republic of Korea; Pohang University of Science and Technology (POSTECH), Republic of Korea
Poster
main
https://github.com/YoungHyun197/ptq4vm
https://openaccess.thecvf.com/content/WACV2025/html/Cho_PTQ4VM_Post-Training_Quantization_for_Visual_Mamba_WACV_2025_paper.html
1
2412.20386
PTQ4VM: Post-Training Quantization for Visual Mamba Visual Mamba is an approach that extends the selective space state model Mamba to vision tasks. It processes image tokens sequentially in a fixed order accumulating information to generate outputs. Despite its growing popularity for delivering high-quality outputs at ...
[ -0.002542586298659444, 0.004986840300261974, -0.012670788913965225, 0.01602344401180744, -0.013494905084371567, 0.02457365207374096, 0.014983934350311756, 0.010544943623244762, 0.023112718015909195, 0.025210468098521233, 0.0034790821373462677, -0.04704955220222473, 0.010011141188442707, 0....
wacv_2025_61fc261c06
61fc261c06
wacv
2,025
PULSE: Physiological Understanding with Liquid Signal Extraction
The non-contact estimation of vital signs particularly heart rate from video data is a promising method for remote health monitoring. 3D convolutional layers are widely used for this task due to their ability to capture both spatial and temporal features. However traditional 3D convolutions while effective in many case...
Shahzad Ahmad; Sania Bano; Sachin Verma; Yogesh Singh Rawat; Sukalpa Chanda; Santosh Kumar Vipparthi; Subrahmanyam Murala
Østfold University College, Norway; Indian Institute of Technology Ropar, India; Norwegian University of Science and Technology, Norway; University of Central Florida, USA; Østfold University College, Norway; Indian Institute of Technology Ropar, India; Trinity College Dublin, Ireland
Poster
main
https://openaccess.thecvf.com/content/WACV2025/html/Ahmad_PULSE_Physiological_Understanding_with_Liquid_Signal_Extraction_WACV_2025_paper.html
0
PULSE: Physiological Understanding with Liquid Signal Extraction The non-contact estimation of vital signs particularly heart rate from video data is a promising method for remote health monitoring. 3D convolutional layers are widely used for this task due to their ability to capture both spatial and temporal features....
[ -0.04142136126756668, -0.003771705087274313, -0.021881932392716408, 0.017466504126787186, -0.0055634393356740475, -0.013292762450873852, 0.0055866786278784275, -0.035992711782455444, 0.01843325048685074, 0.028872251510620117, -0.01726199872791767, 0.007975657470524311, -0.018925918266177177,...
wacv_2025_a8d3d38942
a8d3d38942
wacv
2,025
PV-VTT: A Privacy-Centric Dataset for Mission-Specific Anomaly Detection and Natural Language Interpretation
Video crime detection is a significant application of computer vision and artificial intelligence. However existing datasets primarily focus on detecting severe crimes by analyzing entire video clips often neglecting the precursor activities (i.e. privacy violations) that could potentially prevent these crimes. To addr...
Ryozo Masukawa; Sanggeon Yun; Yoshiki Yamaguchi; Mohsen Imani
University of California, Irvine; University of California, Irvine; Shibaura Institute of Technology; University of California, Irvine
Poster
main
https://openaccess.thecvf.com/content/WACV2025/html/Masukawa_PV-VTT_A_Privacy-Centric_Dataset_for_Mission-Specific_Anomaly_Detection_and_Natural_WACV_2025_paper.html
1
PV-VTT: A Privacy-Centric Dataset for Mission-Specific Anomaly Detection and Natural Language Interpretation Video crime detection is a significant application of computer vision and artificial intelligence. However existing datasets primarily focus on detecting severe crimes by analyzing entire video clips often negle...
[ -0.06632248312234879, -0.04641835391521454, -0.008438021875917912, 0.008036430925130844, -0.03543230518698692, -0.01717146672308445, 0.014771154150366783, 0.016460604965686798, -0.0026634237729012966, 0.005631502717733383, -0.007879487238824368, -0.006545467767864466, -0.018925542011857033, ...
wacv_2025_63e6e4ec3c
63e6e4ec3c
wacv
2,025
PVP: Polar Representation Boost for 3D Semantic Occupancy Prediction
Recently representations based on polar coordinates have exhibited promising characteristics for 3D perceptual tasks. In addition to Cartesian-based methods representing surrounding spaces through polar grids offers a compelling alternative in these tasks. This approach is advantageous for its ability to represent larg...
Yujing Xue; Jiaxiang Liu; Jiawei Du; Joey Tianyi Zhou
NUS; Zhejiang University; A*STAR+NUS; A*STAR
Poster
main
https://openaccess.thecvf.com/content/WACV2025/html/Xue_PVP_Polar_Representation_Boost_for_3D_Semantic_Occupancy_Prediction_WACV_2025_paper.html
0
2412.07616
PVP: Polar Representation Boost for 3D Semantic Occupancy Prediction Recently representations based on polar coordinates have exhibited promising characteristics for 3D perceptual tasks. In addition to Cartesian-based methods representing surrounding spaces through polar grids offers a compelling alternative in these t...
[ -0.04136085510253906, -0.03167508542537689, -0.030197912827134132, 0.01810004934668541, -0.01952112838625908, 0.00036754037137143314, 0.010686133988201618, -0.033806703984737396, 0.05011170729994774, -0.011836085468530655, -0.006277987267822027, 0.017174478620290756, 0.02083001472055912, 0...
wacv_2025_6217274982
6217274982
wacv
2,025
PVT: An Implicit Surface Reconstruction Framework via Point Voxel Geometric-Aware Transformer
3D surface reconstruction from unorganized point clouds is a fundamental task in visual computing with numerous applications in areas such as robotics virtual reality augmented reality and animation. To date many deep learning-based surface reconstruction methods have been proposed demonstrating great performance on ma...
Chuanmao Fan; Chenxi Zhao; Ye Duan
University of Missouri-Columbia; Clemson University; Clemson University
Poster
main
https://openaccess.thecvf.com/content/WACV2025/html/Fan_PVT_An_Implicit_Surface_Reconstruction_Framework_via_Point_Voxel_Geometric-Aware_WACV_2025_paper.html
0
PVT: An Implicit Surface Reconstruction Framework via Point Voxel Geometric-Aware Transformer 3D surface reconstruction from unorganized point clouds is a fundamental task in visual computing with numerous applications in areas such as robotics virtual reality augmented reality and animation. To date many deep learning...
[ -0.02946225181221962, -0.010341963730752468, 0.005006388295441866, 0.003266725456342101, -0.015746118500828743, 0.02251274511218071, 0.03657635301351547, -0.009729309938848019, 0.03491212800145149, 0.0038268007338047028, -0.02710307575762272, -0.0080970898270607, 0.010232234373688698, 0.03...
wacv_2025_3809071d21
3809071d21
wacv
2,025
Paladin: Understanding Video Intentions in Political Advertisement Videos
In this paper we introduce a novel task for video understanding that focuses on detecting editing intentions in political advertisement videos. Political advertisement videos are edited with some intentions (e.g. "associating some candidates with negative emotions") of making people unthinkingly believe the messages in...
Hong Liu; Yuta Nakashima; Noboru Babaguchi
Osaka University, Japan; Osaka University, Japan + Fukui University of Technology, Japan; Osaka University, Japan + Fukui University of Technology, Japan
Poster
main
https://openaccess.thecvf.com/content/WACV2025/html/Liu_Paladin_Understanding_Video_Intentions_in_Political_Advertisement_Videos_WACV_2025_paper.html
0
Paladin: Understanding Video Intentions in Political Advertisement Videos In this paper we introduce a novel task for video understanding that focuses on detecting editing intentions in political advertisement videos. Political advertisement videos are edited with some intentions (e.g. "associating some candidates with...
[ -0.03542974218726158, -0.05316293239593506, 0.017192767933011055, 0.043966587632894516, -0.02046278305351734, 0.013528885319828987, 0.01050618290901184, 0.006965956185013056, -0.017394281923770905, 0.03610755875706673, 0.01343728881329298, -0.045578695833683014, -0.006146162748336792, -0.0...
wacv_2025_e53da2272b
e53da2272b
wacv
2,025
Partial Filter-Sharing: Improved Parameter-Sharing Method for Single Image Super-Resolution Networks
Numerous deep learning techniques have been developed for Single Image Super-Resolution (SISR) leading to significant performance improvements. However these techniques have also resulted in a substantial increase in parameter size. As a result there is a growing interest in reducing network complexity for more practic...
Karam Park; Nam Ik Cho
Department of ECE, INMC, Seoul National University; Department of ECE, INMC, Seoul National University
Poster
main
https://github.com/saturnian77/Partial_filter-Sharing
https://openaccess.thecvf.com/content/WACV2025/html/Park_Partial_Filter-Sharing_Improved_Parameter-Sharing_Method_for_Single_Image_Super-Resolution_Networks_WACV_2025_paper.html
0
Partial Filter-Sharing: Improved Parameter-Sharing Method for Single Image Super-Resolution Networks Numerous deep learning techniques have been developed for Single Image Super-Resolution (SISR) leading to significant performance improvements. However these techniques have also resulted in a substantial increase in pa...
[ -0.011040205135941505, -0.01626547798514366, -0.02761412039399147, 0.03331112116575241, 0.0003350842744112015, -0.015666749328374863, 0.05130928382277489, 0.013217402622103691, -0.010641052387654781, 0.03539760038256645, -0.01694585382938385, 0.01287267915904522, -0.005492887459695339, 0.0...
wacv_2025_bfe559696a
bfe559696a
wacv
2,025
Partial Texture VAE: Color and Texture Encoder for Rock Particle Images
We propose Partial Texture VAE (PT-VAE) for rock particle image analysis a variant of the variational autoencoder (VAE) specialized in encoding color and texture properties from arbitrarily sized and shaped unresized images containing invalid background pixels into the vectors with the same size. PT-VAE integrates part...
Tetsushi Yamada; Simone Di Santo
SLB Schlumberger-Doll Research, Cambridge, Massachusetts, USA; SLB Schlumberger Dhahran Carbonate Research, Dhahran, Saudi Arabia
Poster
main
https://openaccess.thecvf.com/content/WACV2025/html/Yamada_Partial_Texture_VAE_Color_and_Texture_Encoder_for_Rock_Particle_WACV_2025_paper.html
0
Partial Texture VAE: Color and Texture Encoder for Rock Particle Images We propose Partial Texture VAE (PT-VAE) for rock particle image analysis a variant of the variational autoencoder (VAE) specialized in encoding color and texture properties from arbitrarily sized and shaped unresized images containing invalid backg...
[ -0.008680273778736591, -0.016574792563915253, -0.02307344228029251, 0.03767922893166542, 0.0063368771225214005, 0.0001643728610360995, 0.027787968516349792, 0.057091981172561646, 0.0020834042225033045, 0.07081032544374466, 0.009549225680530071, -0.026512272655963898, -0.004453377798199654, ...
wacv_2025_7eec510e07
7eec510e07
wacv
2,025
Patch Ranking: Token Pruning as Ranking Prediction for Efficient CLIP
Contrastive image-text pre-trained models such as CLIP have shown remarkable adaptability to downstream tasks. However they face challenges due to the high computational requirements of the Vision Transformer (ViT) backbone. Current strategies to boost ViT efficiency focus on pruning patch tokens but fall short in addr...
Cheng-En Wu; Jinhong Lin; Yu Hen Hu; Pedro Morgado
University of Wisconsin–Madison; University of Wisconsin–Madison; University of Wisconsin–Madison; University of Wisconsin–Madison
Poster
main
https://github.com/CEWu/PatchRanking
https://openaccess.thecvf.com/content/WACV2025/html/Wu_Patch_Ranking_Token_Pruning_as_Ranking_Prediction_for_Efficient_CLIP_WACV_2025_paper.html
0
Patch Ranking: Token Pruning as Ranking Prediction for Efficient CLIP Contrastive image-text pre-trained models such as CLIP have shown remarkable adaptability to downstream tasks. However they face challenges due to the high computational requirements of the Vision Transformer (ViT) backbone. Current strategies to boo...
[ -0.06571981310844421, -0.015869736671447754, -0.0039949859492480755, 0.0015669987769797444, -0.011902302503585815, -0.022959688678383827, 0.01609015092253685, 0.028047556057572365, -0.0006669789436273277, 0.025549542158842087, -0.015419727191329002, -0.017605489119887352, -0.0124533353373408...
wacv_2025_55fe188aa4
55fe188aa4
wacv
2,025
PatchFinder: Leveraging Visual Language Models for Accurate Information Retrieval using Model Uncertainty
For decades corporations and governments have relied on scanned documents to record vast amounts of information. However extracting this information is a slow and tedious process due to the sheer volume and complexity of these records. The rise of Vision Language Models (VLMs) presents a way to efficiently and accurate...
Roman Colman; Minh Vu; Manish Bhattarai; Martin Ma; Hari Viswanathan; Daniel O'Malley; Javier Santos
;;;;;;
Poster
main
https://openaccess.thecvf.com/content/WACV2025/html/Colman_PatchFinder_Leveraging_Visual_Language_Models_for_Accurate_Information_Retrieval_using_WACV_2025_paper.html
0
2412.02886
PatchFinder: Leveraging Visual Language Models for Accurate Information Retrieval using Model Uncertainty For decades corporations and governments have relied on scanned documents to record vast amounts of information. However extracting this information is a slow and tedious process due to the sheer volume and complex...
[ -0.030255747959017754, -0.04484664648771286, -0.022960297763347626, 0.01743316277861595, -0.004358290694653988, 0.0007794234552420676, 0.03492187336087227, -0.002102765254676342, 0.020664269104599953, 0.0021756733767688274, -0.02238629199564457, -0.014646448194980621, -0.025015614926815033, ...
wacv_2025_1fad4eedbc
1fad4eedbc
wacv
2,025
Pay Attention to Your Neighbours: Training-Free Open-Vocabulary Semantic Segmentation
Despite the significant progress in deep learning for dense visual recognition problems such as semantic segmentation traditional methods are constrained by fixed class sets. Meanwhile vision-language foundation models such as CLIP have showcased remarkable effectiveness in numerous zero-shot image-level tasks owing to...
Sina Hajimiri; Ismail Ben Ayed; Jose Dolz
´ETS Montreal; ´ETS Montreal; ´ETS Montreal
Poster
main
https://github.com/sinahmr/NACLIP
https://openaccess.thecvf.com/content/WACV2025/html/Hajimiri_Pay_Attention_to_Your_Neighbours_Training-Free_Open-Vocabulary_Semantic_Segmentation_WACV_2025_paper.html
20
2404.08181
Pay Attention to Your Neighbours: Training-Free Open-Vocabulary Semantic Segmentation Despite the significant progress in deep learning for dense visual recognition problems such as semantic segmentation traditional methods are constrained by fixed class sets. Meanwhile vision-language foundation models such as CLIP ha...
[ -0.03104480355978012, -0.02635004185140133, -0.0251067653298378, 0.025756238028407097, -0.004945272114127874, -0.008160162717103958, 0.023288240656256676, -0.03616635873913765, 0.02859536185860634, 0.017786279320716858, -0.010437957011163235, -0.0043909004889428616, -0.009143650531768799, ...
wacv_2025_6ca3e07efe
6ca3e07efe
wacv
2,025
Per-Pixel Solution of Multispectral Photometric Stereo
Photometric Stereo (PS) estimates surface normals by analyzing images lit from different angles. Enhancing PS with spectral imaging known as multispectral photometric stereo (MPS) uses varying light source colors for simultaneous image capture. As in traditional PS obtaining a unique solution is challenging in MPS when...
Shin Ishihara; Imari Sato
;
Poster
main
https://openaccess.thecvf.com/content/WACV2025/html/Ishihara_Per-Pixel_Solution_of_Multispectral_Photometric_Stereo_WACV_2025_paper.html
0
Per-Pixel Solution of Multispectral Photometric Stereo Photometric Stereo (PS) estimates surface normals by analyzing images lit from different angles. Enhancing PS with spectral imaging known as multispectral photometric stereo (MPS) uses varying light source colors for simultaneous image capture. As in traditional PS...
[ -0.03984130173921585, -0.011629706248641014, -0.03409639745950699, 0.006514314096421003, -0.005772885400801897, -0.007432939950376749, 0.030571112409234047, -0.004096510354429483, 0.04733952507376671, 0.014045178890228271, -0.051107291132211685, 0.022252188995480537, 0.023781677708029747, ...
wacv_2025_ed57674351
ed57674351
wacv
2,025
Perceive Query & Reason: Enhancing Video QA with Question-Guided Temporal Queries
Video Question Answering (Video QA) is a challenging video understanding task that requires models to comprehend entire videos identify the most relevant information based on contextual cues from a given question and reason accurately to provide answers. Recent advancements in Multimodal Large Language Models (MLLMs) h...
Roberto Amoroso; Gengyuan Zhang; Rajat Koner; Lorenzo Baraldi; Rita Cucchiara; Volker Tresp
LMU Munich+MCML; LMU Munich+MCML; LMU Munich; University of Modena and Reggio Emilia; University of Modena and Reggio Emilia+IIT-CNR; LMU Munich+MCML
Poster
main
https://openaccess.thecvf.com/content/WACV2025/html/Amoroso_Perceive_Query__Reason_Enhancing_Video_QA_with_Question-Guided_Temporal_WACV_2025_paper.html
1
2412.19304
Perceive Query & Reason: Enhancing Video QA with Question-Guided Temporal Queries Video Question Answering (Video QA) is a challenging video understanding task that requires models to comprehend entire videos identify the most relevant information based on contextual cues from a given question and reason accurately to ...
[ -0.04585445672273636, -0.01134636253118515, -0.01094049122184515, -0.005064374301582575, -0.036510393023490906, 0.043401189148426056, -0.021682556718587875, 0.013718456029891968, -0.007522151339799166, 0.0032740298192948103, 0.02438836544752121, -0.06006897985935211, -0.01937359943985939, ...
wacv_2025_89e6d0e916
89e6d0e916
wacv
2,025
Personalized Mixture of Experts for Multi-Site Medical Image Segmentation
The sharing of sensitive medical data among institutions presents a significant challenge due to strict privacy regulations the need for robust de-identification processes and the ethical imperative to protect patient confidentiality. Federated Learning (FL) addresses these challenges by enabling institutions to collab...
Md Motiur Rahman; Mohamed Trabelsi; Huseyin Uzunalioglu; Aidan Boyd
Nokia Bell Labs; Nokia Bell Labs; Nokia Bell Labs; Nokia Bell Labs
Poster
main
https://openaccess.thecvf.com/content/WACV2025/html/Rahman_Personalized_Mixture_of_Experts_for_Multi-Site_Medical_Image_Segmentation_WACV_2025_paper.html
0
Personalized Mixture of Experts for Multi-Site Medical Image Segmentation The sharing of sensitive medical data among institutions presents a significant challenge due to strict privacy regulations the need for robust de-identification processes and the ethical imperative to protect patient confidentiality. Federated L...
[ -0.014805816113948822, -0.05508022382855415, -0.02339707314968109, -0.0008109212503768504, 0.006894276011735201, 0.004616991616785526, 0.04771893098950386, -0.02750311978161335, -0.0048551238141953945, 0.013270671479403973, -0.02088165655732155, 0.012706552632153034, -0.011707784608006477, ...
wacv_2025_4978a57c82
4978a57c82
wacv
2,025
Phaseformer: Phase-Based Attention Mechanism for Underwater Image Restoration and Beyond
Quality degradation is observed in underwater images due to the effects of light refraction and absorption by water leading to issues like color cast haziness and limited visibility. This degradation negatively affects the performance of autonomous underwater vehicles used in marine applications. To address these chall...
Raqib Khan; Anshul Negi; Ashutosh Kulkarni; Shruti S. Phutke; Santosh Kumar Vipparthi; Subrahmanyam Murala
Trinity College Dublin, Ireland; Indian Institute of Technology Ropar, India; Indian Institute of Technology Ropar, India; Indian Institute of Technology Ropar, India; Indian Institute of Technology Ropar, India; Trinity College Dublin, Ireland
Poster
main
https://openaccess.thecvf.com/content/WACV2025/html/Khan_Phaseformer_Phase-Based_Attention_Mechanism_for_Underwater_Image_Restoration_and_Beyond_WACV_2025_paper.html
1
2412.01456
Phaseformer: Phase-Based Attention Mechanism for Underwater Image Restoration and Beyond Quality degradation is observed in underwater images due to the effects of light refraction and absorption by water leading to issues like color cast haziness and limited visibility. This degradation negatively affects the performa...
[ -0.04314906522631645, -0.016669537872076035, -0.038495369255542755, -0.01268597412854433, 0.005067874677479267, 0.01124332845211029, 0.02418060228228569, 0.039388880133628845, -0.013402643613517284, 0.02311955951154232, -0.02358493022620678, 0.01505935937166214, -0.03758324682712555, 0.026...
wacv_2025_6db3fa4366
6db3fa4366
wacv
2,025
Physiology-Aware PolySnake for Coronary Vessel Segmentation
Coronary artery disease (CAD) is a significant health risk that requires early detection for effective treatment. While recent advances in deep learning have shown promise in automating CAD detection from coronary computed tomography angiography (CCTA) images the accurate segmentation of coronary vessels remains a chal...
Yizhe Ruan; Lin Gu; Yusuke Kurose; Junichi Iho; Youji Tokunaga; Makoto Horie; Yusaku Hayashi; Keisuke Nishizawa; Yasushi Koyama; Tatsuya Harada
The University of Tokyo + RIKEN Center for Advanced Intelligence Project; RIKEN Center for Advanced Intelligence Project + The University of Tokyo; The University of Tokyo + RIKEN Center for Advanced Intelligence Project; Sakurabashi Watanabe Advanced Healthcare Hospital; Sakurabashi Watanabe Advanced Healthcare Hospit...
Poster
main
https://github.com/opensourcetorch/Physiology-aware-PolySnake
https://openaccess.thecvf.com/content/WACV2025/html/Ruan_Physiology-Aware_PolySnake_for_Coronary_Vessel_Segmentation_WACV_2025_paper.html
0
Physiology-Aware PolySnake for Coronary Vessel Segmentation Coronary artery disease (CAD) is a significant health risk that requires early detection for effective treatment. While recent advances in deep learning have shown promise in automating CAD detection from coronary computed tomography angiography (CCTA) images ...
[ -0.01901177316904068, -0.013763879425823689, -0.042399950325489044, 0.010988369584083557, 0.006214112043380737, 0.003263356164097786, 0.01918228156864643, 0.0022130669094622135, -0.025595322251319885, 0.05494184419512749, 0.028588704764842987, -0.008421259000897408, -0.027035176753997803, ...
wacv_2025_f6a15d5df3
f6a15d5df3
wacv
2,025
PivotAlign: Improve Semi-Supervised Learning by Learning Intra-Class Heterogeneity and Aligning with Pivots
Self-supervised learning plays an important role in current state-of-the-art semi-supervised learning (SSL) methods. These methods learn inter-class heterogeneity among data and generate pseudo-labels based on class level representations. However they often neglect intra-class heterogeneity resulting in the under-explo...
Lingjie Yi; Tao Sun; Yikai Zhang; Songzhu Zheng; Weimin Lyu; Haibin Ling; Chao Chen
Stony Brook University; Stony Brook University; Morgan Stanley; Morgan Stanley; Stony Brook University; Stony Brook University; Stony Brook University
Poster
main
https://openaccess.thecvf.com/content/WACV2025/html/Yi_PivotAlign_Improve_Semi-Supervised_Learning_by_Learning_Intra-Class_Heterogeneity_and_Aligning_WACV_2025_paper.html
0
PivotAlign: Improve Semi-Supervised Learning by Learning Intra-Class Heterogeneity and Aligning with Pivots Self-supervised learning plays an important role in current state-of-the-art semi-supervised learning (SSL) methods. These methods learn inter-class heterogeneity among data and generate pseudo-labels based on cl...
[ -0.06778687238693237, -0.02238141931593418, -0.023226000368595123, 0.0054851919412612915, 0.011539558880031109, 0.038410115987062454, 0.03954846411943436, 0.03137805312871933, -0.005645846016705036, 0.0358029268682003, -0.008826798759400845, -0.006481247954070568, -0.02814660780131817, 0.0...
wacv_2025_c1df1e88f9
c1df1e88f9
wacv
2,025
Pix2Poly: A Sequence Prediction Method for End-to-End Polygonal Building Footprint Extraction from Remote Sensing Imagery
Extraction of building footprint polygons from remotely sensed data is essential for several urban understanding tasks such as reconstruction navigation & mapping. Despite significant progress in the area extracting accurate polygonal vector building footprints remains an open problem. In this paper we introduce Pix2Po...
Yeshwanth Kumar Adimoolam; Charalambos Poullis; Melinos Averkiou
CYENS CoE, Cyprus; Concordia University; CYENS CoE, Cyprus
Poster
main
https://github.com/yeshwanth95/Pix2Poly
https://openaccess.thecvf.com/content/WACV2025/html/Adimoolam_Pix2Poly_A_Sequence_Prediction_Method_for_End-to-End_Polygonal_Building_Footprint_WACV_2025_paper.html
1
2412.07899
Pix2Poly: A Sequence Prediction Method for End-to-End Polygonal Building Footprint Extraction from Remote Sensing Imagery Extraction of building footprint polygons from remotely sensed data is essential for several urban understanding tasks such as reconstruction navigation & mapping. Despite significant progress in th...
[ -0.05780757963657379, -0.008786525577306747, -0.011065914295613766, 0.04899267852306366, -0.010564638301730156, -0.0447554737329483, 0.011179410852491856, -0.013903328217566013, 0.010139025747776031, 0.00391799584031105, -0.029452357441186905, 0.041842397302389145, -0.0008435394847765565, ...
wacv_2025_7c8092cf8a
7c8092cf8a
wacv
2,025
PixSwap: High-Resolution Face Swapping for Effective Reflection of Identity via Pixel-Level Supervision with Synthetic Paired Dataset
Face swapping is to interchange identity features such as eyes nose and lips between a source and a target face while preserving the target attributes like expression pose skin color and hair. Despite considerable advancements in quality over the years recent studies conducting high-resolution face swapping still encou...
Taewoo Kim; Geonsu Lee; Hyukgi Lee; Seongtae Kim; Younggun Lee
Neosapience; Neosapience; Neosapience; Neosapience; Neosapience
Poster
main
https://openaccess.thecvf.com/content/WACV2025/html/Kim_PixSwap_High-Resolution_Face_Swapping_for_Effective_Reflection_of_Identity_via_WACV_2025_paper.html
0
PixSwap: High-Resolution Face Swapping for Effective Reflection of Identity via Pixel-Level Supervision with Synthetic Paired Dataset Face swapping is to interchange identity features such as eyes nose and lips between a source and a target face while preserving the target attributes like expression pose skin color and...
[ -0.06200474873185158, -0.045271556824445724, -0.028869351372122765, -0.012439564801752567, -0.01264183409512043, -0.013000402599573135, -0.000387012492865324, 0.019142035394906998, 0.01531730592250824, -0.000022410524252336472, -0.010177825577557087, -0.0027191436383873224, -0.01745032705366...
wacv_2025_63ad554933
63ad554933
wacv
2,025
Pixel-Wise Shuffling with Collaborative Sparsity for Melanoma Hyperspectral Image Classification
Hyperspectral imaging has emerged as a promising technology for medical image classification particularly in skin cancer diagnosis. However current methods face significant challenges in accurately and robustly classifying non-cancerous skin lesions especially when melanoma lesions overlap with pigmented regions. Exist...
Favour Ekong; Jun Zhou; Kwabena Sarpong; Yongsheng Gao
Griffith University, Queensland, Australia; Griffith University, Queensland, Australia; Griffith University, Queensland, Australia; Griffith University, Queensland, Australia
Poster
main
https://openaccess.thecvf.com/content/WACV2025/html/Ekong_Pixel-Wise_Shuffling_with_Collaborative_Sparsity_for_Melanoma_Hyperspectral_Image_Classification_WACV_2025_paper.html
0
Pixel-Wise Shuffling with Collaborative Sparsity for Melanoma Hyperspectral Image Classification Hyperspectral imaging has emerged as a promising technology for medical image classification particularly in skin cancer diagnosis. However current methods face significant challenges in accurately and robustly classifying ...
[ -0.09061940759420395, -0.026512544602155685, -0.03539251908659935, 0.01339274737983942, -0.0022325036115944386, 0.013993238098919392, -0.0038849886041134596, 0.0018344515701755881, 0.020052727311849594, 0.038467757403850555, -0.019634203985333443, 0.010535870678722858, 0.024019600823521614, ...