CN-CVS: A Mandarin Audio-Visual Dataset for Large Vocabulary Continuous Visual to Speech Synthesis.

Chen Chen, Dong Wang, Thomas Fang Zheng

Browse the full ICASSP paper archive.