The-performance-optimization-of-Wav2Lip

The wav2lip algorithm has shown excellent performance in scientific data, but it cannot handle the speech containing noise in the wild world very well. This project combines the DTLN to first perform noise reduction processing on the speech, and then use the FILTER to carry out Filter processing can achieve better performance. The execution steps are:

1. DTLN
2. FILTER
3. Wav2Lip