IA-NET: Acceleration and Compression of Speech Enhancement Using Integer-Adder Deep Neural Network.
Yu-Chen LinYi-Te HsuSzu-Wei FuYu TsaoTei-Wei KuoPublished in: INTERSPEECH (2019)
Keyphrases
- speech enhancement
- neural network
- noisy environments
- noise reduction
- single channel
- signal to noise ratio
- linear prediction
- speech signal
- compression algorithm
- pattern recognition
- lossless compression
- vocal tract
- image compression
- compression ratio
- background noise
- smoothing algorithm
- multi channel
- edge detection
- random fields
- edge preserving
- additive noise
- optical flow