Login / Signup

Transcribe-to-Diarize: Neural Speaker Diarization for Unlimited Number of Speakers Using End-to-End Speaker-Attributed ASR.

Naoyuki KandaXiong XiaoYashesh GaurXiaofei WangZhong MengZhuo ChenTakuya Yoshioka
Published in: ICASSP (2022)
Keyphrases
  • end to end
  • speaker diarization
  • speech recognition
  • automatic speech recognition
  • computational complexity
  • maximum likelihood
  • non stationary
  • network architecture