Streaming Speaker-Attributed ASR with Token-Level Speaker Embeddings.
Naoyuki KandaJian WuYu WuXiong XiaoZhong MengXiaofei WangYashesh GaurZhuo ChenJinyu LiTakuya YoshiokaPublished in: INTERSPEECH (2022)
Keyphrases
- automatic speech recognition
- speech recognition
- speaker verification
- audio visual
- speaker recognition
- speaker identification
- speaker diarization
- noisy environments
- euclidean space
- levels of abstraction
- data streams
- acoustic features
- language model
- broadcast news
- speech signal
- learning algorithm
- low dimensional
- distance measure
- hidden markov models
- pattern recognition