Time-Domain Speech Enhancement for Robust Automatic Speech Recognition.
Yufeng YangAshutosh PandeyDeLiang WangPublished in: INTERSPEECH (2023)
Keyphrases
- noisy environments
- automatic speech recognition
- speech enhancement
- speech signal
- speech recognition
- hidden markov models
- speaker verification
- noise reduction
- background noise
- speaker identification
- word error rate
- vocal tract
- conversational speech
- linear prediction
- speech retrieval
- broadcast news
- spectral subtraction
- additive noise
- signal to noise ratio
- non stationary
- neural network
- frequency domain
- multi modal
- high resolution
- multiresolution
- image processing
- computer vision