---
title: SA-IQA Spatial Aesthetics
emoji: 🏠
colorFrom: red
colorTo: purple
sdk: gradio
sdk_version: 6.25.0
app_file: app.py
short_description: Assess spatial aesthetics of interior images
python_version: "3.12"
startup_duration_timeout: 30m
---

# SA-IQA: Spatial Aesthetics Image Quality Assessment

This Space demonstrates [SA-IQA](https://huggingface.co/gaoyuan-ai/SA-IQA-model), a multimodal
vision-language model for assessing the spatial aesthetic quality of interior images.

The model is a fine-tuned [Ovis2.5-9B](https://huggingface.co/AIDC-AI/Ovis2.5-9B) that evaluates
interior images across four dimensions: **distortion**, **harmony**, **layout**, and **lighting**.

## Paper

"Beyond Pixels: Benchmarking and Reward-Based Assessing Framework for Visual Spatial Aesthetics" (CVPRW 2026).

- [arXiv](https://arxiv.org/abs/2512.05098)
- [GitHub](https://github.com/AlibabaResearch/SA-IQA)
- [Dataset (SA-BENCH)](https://huggingface.co/datasets/gaoyuan-ai/SA-BENCH)

## Model

The SA-IQA model (`sa-iqa-prompt4` checkpoint) is loaded from the nested subdirectory
`gaoyuan-ai/SA-IQA-model/sa-iqa-prompt4` as recommended by the authors.

## License

Apache 2.0. Example images are from [SA-BENCH](https://huggingface.co/datasets/gaoyuan-ai/SA-BENCH) (Apache 2.0).