| --- |
| license: cc-by-nc-4.0 |
| pipeline_tag: video-text-to-text |
| --- |
| |
| # ST-Evidence-7B |
|
|
| This repository contains the official **ST-Evidence-7B** model introduced in the paper [Evidence-Backed Video Question Answering](https://huggingface.co/papers/2607.11862). |
|
|
| * **Code**: [GitHub - SalesforceAIResearch/EVQA](https://github.com/SalesforceAIResearch/EVQA) |
| * **Benchmark (ST-Evidence)**: [Salesforce/ST-Evidence-Bench](https://huggingface.co/datasets/Salesforce/ST-Evidence-Bench) |
| * **SFT Dataset (ST-Evidence-Instruct)**: [Salesforce/ST-Evidence-Instruct](https://huggingface.co/datasets/Salesforce/ST-Evidence-Instruct) |
|
|
| ## Model Description |
|
|
| We propose **Evidence-Backed Video Question Answering** (E-VQA), a task where multimodal models are designed to jointly produce a semantic **textual answer** and associated spatiotemporal evidence. This evidence includes **temporal segments** and **dense, tracked object segmentation masklets**. A masklet is defined as a temporal sequence of object segmentation masks tracked over time. |
|
|
| Our model was fine-tuned from UniPixel, which is built upon Qwen2.5-VL and SAM 2.1. UniPixel is a unified model designed to handle both video question answering and mask generation. |
|
|
| This model was released for research purposes only, in support of the academic paper *Evidence-Backed Video Question Answering*. |
|
|
| ## Citation |
|
|
| If you find this work useful for your research, please cite our paper: |
|
|
| ```bibtex |
| @inproceedings{wang2026evidence, |
| title={Evidence-Backed Video Question Answering}, |
| author={Wang, Shijie and Zhou, Honglu and Wang, Ziyang and Xu, Ran and Xiong, Caiming and Savarese, Silvio and Sun, Chen and Niebles, Juan Carlos}, |
| booktitle={European Conference on Computer Vision (ECCV)}, |
| year={2026} |
| } |
| ``` |
