tubbz-alt / cocktail-party

Official Implementation of "Seeing Through Noise: Speaker Separation and Enhancement using Visually-derived Speech", ICASSP 2018.

Seeing Through Noise: Speaker Separation and Enhancement using Visually-derived Speech

Implementation of the methods described in the paper: Seeing Through Noise: Speaker Separation and Enhancement using Visually-derived Speech by Aviv Gabbay, Ariel Ephrat, Tavi Halperin and Shmuel Peleg.

Speaker Separation and Enhancement Demo

Usage

Dependencies

• python >= 2.7
• mediaio
• face-detection
• keras >= 2.0.4
• numpy >= 1.12.1
• dlib >= 19.4.0
• opencv >= 3.2.0

Citing

If you find this project useful for your research, please cite

@inproceedings{gabbay2018seeing,
  author    = {Aviv Gabbay and
               Ariel Ephrat and
               Tavi Halperin and
               Shmuel Peleg},
  title     = {Seeing Through Noise: Visually Driven Speaker Separation And Enhancement},
  booktitle = {{ICASSP}},
  pages     = {3051--3055},
  publisher = {{IEEE}},
  year      = {2018}
}