Segmentation of Singing, Speech and Instruments in Kunqu Audio Based on Zero-Crossing Rate.
Gen-Fang ChenYa-Dong WuPublished in: ISCID (1) (2019)
Keyphrases
- zero crossing
- audio features
- audio visual
- edge detection
- audio stream
- acoustic features
- level set
- broadcast news
- segmentation method
- text to speech
- speaker identification
- music information retrieval
- segmentation algorithm
- image segmentation
- emotion recognition
- edge detector
- scale space
- medical images
- speech signal
- speech recognition
- shape prior
- image intensity
- feature set
- multiscale
- speaker verification
- gaussian mixture model
- region growing
- directional derivatives
- object recognition
- wavelet transform
- computer vision
- face recognition
- high quality
- video sequences
- image analysis
- visual features