|
Download README.md from DL3DV/DL3DV-10K-Sample: direct link, hf CLI and curl.
- Browser
- Download file 1.74 kB
-
https://huggingface.co/datasets/DL3DV/DL3DV-10K-Sample/resolve/main/README.md
- Command line
-
hf download hf://datasets/DL3DV/DL3DV-10K-Sample/README.md
-
curl -L -H "Authorization: Bearer $HF_TOKEN" -o README.md https://huggingface.co/datasets/DL3DV/DL3DV-10K-Sample/resolve/main/README.md
1.74 kB
metadata
tags:
- novel view synthesis
- NeRF
- 3D Gaussian Splatting
- 3D Vision
- Content Generation
- text-to-3d
- image-to-3d
size_categories:
- 100B<n<1T
DL3DV-10K Dataset Sample
Here is the repo for a sample (11 scenes/videos) of the full DL3DV-10K dataset. Each scene has a raw 4K video and the processed colmap results.
File Structure Example
5c3af581028068a3c402c7cbe16ecf9471ddf2897c34ab634b7b1b6cf81aba00
├── colmap
│ ├── colmap
│ │ ├── features.h5
│ │ ├── global-feats-netvlad.h5
│ │ ├── matches.h5
│ │ ├── pairs-netvlad.txt
│ │ └── sparse
│ │ └── 0
│ │ ├── cameras.bin
│ │ ├── database.db
│ │ ├── images.bin
│ │ ├── models
│ │ │ └── 0
│ │ └── points3D.bin
│ ├── images
│ │ ├── frame_00001.png
│ │ ├── frame_00002.png
│ │ ├── ....
│ ├── images_2
│ │ ├── frame_00001.png
│ │ ├── frame_00002.png
│ │ ├── ....
│ ├── images_4
│ │ ├── frame_00001.png
│ │ ├── frame_00002.png
│ │ ├── ....
│ ├── images_8
│ │ ├── frame_00001.png
│ │ ├── frame_00002.png
│ │ ├── ....
│ ├── temp_images
│ └── transforms.json
└── video.mp4
DL3DV-10K Benchmark
Please refer to this repo for instruction.