Skip to content

ACM Multimedia 2023 (Oral) - RTQ: Rethinking Video-language Understanding Based on Image-text Model

License

Notifications You must be signed in to change notification settings

SCZwangxiao/RTQ-MM2023

Repository files navigation

✨ Highlights

With a simple yet insightful framework RTQ (Refine, Temporal model, and Query), our model demonstrates outstanding performance even in the absence of video-language pre-training.

🏆 Contributions

  1. Our systemic analysis reveals that current methods focus only on restricted aspects of video-language understanding, and they are complementary.

  1. We propose the RTQ framework to jointly model information redundancy, temporal dependency, and scene complexity in video-language understanding.

  1. We demonstrate that, even without pre-training on video-languag data, our method can achieve superior (or comparable) performance with state-of-the-art pre-training methods.

📊 Main Results

Text-to-video retrieval

Video caption

Video question answering

🔧 Requirements and installation

  • Python >= 3.8
  • Pytorch >= 1.10.0
  • CUDA Version >= 10.2
  • Install required packages:
pip install -r requirements.txt

🚢 Download datasets and pretrained models

Follow the instructions in [REPO_HOME]/tools/data to download all datasets. Put them in the [REPO_HOME]/data directory. You can use softlinks as well.

Download BLIP model

mkdir [REPO_HOME]/modelzoo
wget https://storage.googleapis.com/sfr-vision-language-research/BLIP/models/model_base_capfilt_large.pth -P [REPO_HOME]/modelzoo/BLIP

The final file structure is:

- RTQ
  - assets
  - configs
  - data
    - msrvtt
      - txt_db
      - vis_db
    - nextqa
    ......
  - lavis
  - modelzoo
    - BLIP
      - model_base_capfilt_large.pth
  ......

🚀 Training & evaluation

See code examples.

📖 Citation

If you find our paper and code useful in your research, please consider giving a star ⭐ and citation 📖.

@inproceedings{wang2023rtq,
  author       = {Xiao Wang and
                  Yaoyu Li and
                  Tian Gan and
                  Zheng Zhang and
                  Jingjing Lv and
                  Liqiang Nie},
  title        = {{RTQ:} Rethinking Video-language Understanding Based on Image-text
                  Model},
  booktitle    = {Proceedings of the {ACM} International Conference on Multimedia, 2023},
  pages        = {557--566},
  publisher    = {{ACM}},
  year         = {2023},
}

🖖 Acknowledgements

  • LAVIS: great library in vision-language understanding.
  • mmaction2: great toolbox for video understanding.

Releases

No releases published

Packages

No packages published