Visual Speech Recognition (VSR) is the task of predicting spoken words from silent lip
movements. VSR is regarded as a challenging task because of the insufficient information …

We present a novel resource-efficient end-to-end architecture for lipreading that achieves
state-of-the-art results on a popular and challenging benchmark. In particular, we make the …

With the continuous development of deep learning, significant progress has been made in
the field of lip recognition, and many deep learning algorithms for lip recognition have …