Large-Scale Visual Speech Recognition

Benchmark Model Rank Results
lipreading-on-lrs3-tedCTC-V2P–Word Error Rate (WER): 55.1