Wav2code: Restore Clean Speech Representations via Codebook Lookup for Noise-Robust ASR.
Yuchen HuChen ChenQiushi ZhuEng Siong ChngPublished in: CoRR (2023)
Keyphrases
- noisy environments
- speech recognition
- automatic speech recognition
- noisy speech
- speech signal
- speech enhancement
- background noise
- noise reduction
- image noise
- bag of words
- vector quantization
- noisy data
- signal to noise ratio
- salt pepper
- additive noise
- source code
- geometric distortions
- vector quantizer
- broadcast news
- spoken language
- spoken words
- random noise
- image retrieval
- estimation error
- computationally efficient
- language model
- feature vectors