README.md
June 24, 2026 ยท View on GitHub
[ECCV 2026] The Implementation of Em-Garde: A proposal-match framework for proactive streaming video understanding

Installation
Basic environment:
conda create -n em_garde python==3.10
conda activate em_garde
pip install -r requirements.txt
This project requires ffmpeg for video processing. Install it via your system package manager:
# Ubuntu/Debian
sudo apt install ffmpeg
# macOS
brew install ffmpeg
# Conda
conda install -c conda-forge ffmpeg
Usage
run demo:
python -m demo --yaml-path=configs/demo/demo_solution
The video and query can be customized in configs/demo. Refer to configs/demo/demo_solution.yaml. The triggering results will be stored under demo/.
run proactive evaluation:
prepare StreamingBench, OVO-Bench and ProactiveVideoQA data according to official repo. Note that we use src_videos instead of chunked_videos for OVO-Bench.
StreamingBench Evaluation:
python -m eval.streamingbench.streamingbench_eval_online --data-path=/path/to/StreamingBench/src/data/questions_proactive.json
OVO-Bench Evaluation:
python -m eval.ovo_bench.ovo_bench_eval_online --data-path=/path/to/OVO-Bench/data/ovo_bench_new.json --video-root=/path/to/OVO-Bench/data/src_videos
ProactiveVideoQA Evaluation (EGO):
python -m eval.proactive_video_qa.eval_proactive_video_qa --data-path=/path/to/proactive_video_qa/EGO/anno.json
Please follow ProactiveVideoQA to compute the PAUC metric after obtaining the results.
Acknowledgements
This codebase includes adapted code from the following public projects:
vlm2vec/is based on VLM2Vec.embedding_models/ops_mm_embedding_v1.pyis adapted from Ops-MM-Embedding.
We thank the original authors for making their code publicly available.
Cite
If you find our project useful, please cite:
@misc{zheng2026emgarde,
title={Em-Garde: A Propose-Match Framework for Proactive Streaming Video Understanding},
author={Yikai Zheng and Xin Ding and Yifan Yang and Shiqi Jiang and Hao Wu and Qianxi Zhang and Weijun Wang and Ting Cao and Yunxin Liu},
year={2026},
eprint={2603.19054},
archivePrefix={arXiv},
primaryClass={cs.CV},
url={https://arxiv.org/abs/2603.19054},
}