Variable-Length Speaker Conditioning in Flow-Based Text-to-Speech.
Byoung Jin ChoiMyeonghun JeongMinchan KimNam Soo KimPublished in: IEEE Signal Process. Lett. (2024)
Keyphrases
- variable length
- text to speech
- prosodic features
- fixed length
- speech synthesis
- n gram
- speaker verification
- word processing
- speech recognition
- programming tool
- statistical dependencies
- text compression
- bitstream
- text to speech synthesis
- human motion
- spontaneous speech
- audio visual
- similarity measure
- convolutional codes
- image sequences