GAN based Lip Movement synthesis [https://github.com/ayushanand18/aud2lips]
- TensorFlow Keras based conditional Generative Adversarial Network model to synthesize lip movement. The model takes input as a facial image and spectrogram of the speech and generates a facial image with corresponding lip movement.
- Currently under development.
- Make sure you have Jupyter, TensorFlow Keras, matplotlib, opencv, numpy, tqdm and other requirement libraries installed.
- Run the
2-train-gan-relook.ipynb
Notebook.
Ayush Anand
@inproceedings{RekikICIAR14,
author = {Ahmed Rekik and Achraf {Ben-Hamadou} and Walid Mahdi},
title = {A New Visual Speech Recognition Approach for {RGB-D} Cameras},
booktitle = {Image Analysis and Recognition - 11th International Conference, {ICIAR} 2014, Vilamoura, Portugal, October 22-24, 2014}
year = {2014}, pages = {21--28} }
@misc{https://doi.org/10.48550/arxiv.1611.07004,
doi = {10.48550/ARXIV.1611.07004},
url = {https://arxiv.org/abs/1611.07004},
author = {Isola, Phillip and Zhu, Jun-Yan and Zhou, Tinghui and Efros, Alexei A.},
keywords = {Computer Vision and Pattern Recognition (cs.CV), FOS: Computer and information sciences, FOS: Computer and information sciences},
title = {Image-to-Image Translation with Conditional Adversarial Networks},
publisher = {arXiv},
year = {2016},
copyright = {arXiv.org perpetual, non-exclusive license}
}