Beyond Voice Activity Detection: Hybrid Audio Segmentation for Direct Speech Translation.
Marco GaidoMatteo NegriMauro CettoloMarco TurchiPublished in: ICNLSP (2021)
Keyphrases
- voice activity detection
- noisy environments
- speech recognition
- segmentation algorithm
- segmentation method
- multiscale
- image segmentation
- speech signal
- level set
- cross language information retrieval
- medical images
- edge detection
- speaker identification
- speaker verification
- cross language
- maximum likelihood
- word segmentation
- audio features
- text to speech
- audio signals
- speech processing
- digital audio
- feature extraction