SEQ-CPC : Sequential Contrastive Predictive Coding for Automatic Speech Recognition.
Yulong ChenJianping ZhaoWeiqi WangMing FangHaimei KangLu WangTao WeiJun MaShaojun WangJing XiaoPublished in: ICASSP (2021)
Keyphrases
- automatic speech recognition
- predictive coding
- speech recognition
- lossless compression
- speech signal
- motion compensated
- video coding
- binary tree
- hidden markov models
- broadcast news
- concept drift
- entropy coding
- speech retrieval
- arithmetic coding
- conversational speech
- non stationary
- neural network
- data structure
- distributed video coding
- transform coding
- speech sounds
- low complexity
- bit rate
- image quality
- motion estimation
- bayesian networks
- image processing
- computer vision