1 paper · 1 filter
Mohan Shi, Zengrui Jin, Yaoxun Xu +6
Recognizing overlapping speech from multiple speakers in conversational scenarios is one of the most challenging problem for automatic speech recognition (ASR). Serialized output t…