A Speech-Noise-Equilibrium Loss Function for Deep Learning-Based Speech Enhancement.
Weitong ZhaoFushi XieKang OuyangNengheng ZhengPublished in: ISCSLP (2022)
Keyphrases
- speech enhancement
- loss function
- deep learning
- noisy environments
- noise reduction
- signal to noise ratio
- speech signal
- single channel
- noisy speech
- vocal tract
- pairwise
- linear prediction
- unsupervised learning
- additive noise
- speech recognition
- support vector
- machine learning
- background noise
- mental models
- wiener filter
- weakly supervised
- speech synthesis
- automatic speech recognition
- computer vision
- noisy images
- prior information
- image processing