PSST! Prosodic Speech Segmentation with Transformers.
Nathan RollCalbert GrahamSimon ToddPublished in: CoNLL (2023)
Keyphrases
- speech recognition
- text to speech synthesis
- text to speech
- segmentation method
- fully automatic
- segmentation algorithm
- image segmentation
- level set
- prosodic features
- speech synthesis
- region growing
- shape prior
- neural network
- multiscale
- synthesized speech
- object segmentation
- audio visual
- recognition engine
- automatic speech recognition
- pattern recognition
- brain mri
- image analysis
- segmentation errors
- word segmentation
- information retrieval
- optimal segmentation
- image regions
- segmentation accuracy
- speech signal
- motion segmentation
- optical flow
- hidden markov models
- medical images