On the Interplay between Sparsity, Naturalness, Intelligibility, and Prosody in Speech Synthesis.
Cheng-I Jeff LaiErica CooperYang ZhangShiyu ChangKaizhi QianYi-Lun LiaoYung-Sung ChuangAlexander H. LiuJunichi YamagishiDavid D. CoxJames R. GlassPublished in: ICASSP (2022)
Keyphrases
- speech synthesis
- speech recognition
- text to speech
- prosodic features
- vocal tract
- hidden markov models
- language model
- pattern recognition
- sparse representation
- speech signal
- high dimensional
- automatic speech recognition
- sparsity constraints
- computer vision
- multiscale
- mixed norm
- dimensionality reduction
- video sequences
- feature extraction
- case study
- image processing