---

Deep VSR (speechreading) with Face Inputs

Description

We provide code and models (in PyTorch) that can be used to evaluate the methods in our paper Can We Read Speech Beyond the Lips? Rethinking RoI Selection for Deep Visual Speech Recognition. We provide models trained on the LRW dataset (English) and the LRW-1000 dataset (Mandarin Chinese).

Content

• Preprocessing

• How to test

• Model Zoo

• Citation

• License

• Contact

Model Zoo

Coming soon...

Citation

@inproceedings{zhang2020can,
    author = {Y. Zhang and S. Yang and J. Xiao and S. Shan and X. Chen},
    booktitle = {2020 15th IEEE International Conference on Automatic Face and Gesture Recognition (FG 2020) (FG)},
    title = {Can We Read Speech Beyond the Lips? {R}ethinking {RoI} Selection for Deep Visual Speech Recognition},
    year = {2020},
    pages = {851-858},
    keywords = {visual speech recognition},
    doi = {10.1109/FG47880.2020.00134},
    url = {https://doi.ieeecomputersociety.org/10.1109/FG47880.2020.00134},
    publisher = {IEEE Computer Society},
    address = {Los Alamitos, CA, USA}
}

License

TBD

Contact

Yuanhang Zhang