Attention-Based Multi-Task Learning for Speech-Enhancement and Speaker-Identification in Multi-Speaker Dialogue Scenario.
Chiang-Jen PengYun-Ju ChanCheng YuSyu-Siang WangYu TsaoTai-Shih ChiPublished in: ISCAS (2021)
Keyphrases
- speaker identification
- multi task learning
- noisy environments
- speech signal
- multi task
- speech recognition
- speaker recognition
- learning tasks
- vocal tract
- gaussian mixture model
- automatic speech recognition
- noise reduction
- mel frequency cepstral coefficients
- transfer learning
- broadcast news
- feature extraction
- learning algorithm
- neural network
- text mining
- hidden markov models
- collaborative filtering
- multi class
- information extraction
- prior knowledge
- speaker diarization
- machine learning