Lightweight U-Net Based Monaural Speech Source Separation for Edge Computing Device.
Kwang Myung JeonChanjun ChunGyeongbong KimChaejun LeemBohee KimWooyeol ChoiPublished in: ICCE (2020)
Keyphrases
- lightweight
- source separation
- blind source separation
- audio features
- rfid tags
- speech recognition
- speech signal
- independent component analysis
- audio visual
- single channel
- temporal structure
- edge detection
- principal component analysis
- denoising
- sparse representation
- image processing
- low level
- pattern recognition
- music information retrieval
- multiscale
- multimedia