Utterance-level Permutation Invariant Training with Latency-controlled BLSTM for Single-channel Multi-talker Speech Separation.
Lu HuangGaofeng ChengPengyuan ZhangYi YangShumin XuJiasong SunPublished in: APSIPA (2019)
Keyphrases
- single channel
- sound source
- speech recognition
- speech enhancement
- speech signal
- multi channel
- spoken language
- audio visual
- speech synthesis
- independent component analysis
- pattern recognition
- non stationary
- noise reduction
- computer vision
- hidden markov models
- automatic speech recognition
- linear prediction
- training data
- image processing