Investigation of a Single-Channel Frequency-Domain Speech Enhancement Network to Improve End-to-End Bengali Automatic Speech Recognition Under Unseen Noisy Conditions.
Md Mahbub E. NoorYen-Ju LuSyu-Siang WangSupratip GhoseChia-Yu ChangRyandhimas E. ZezarioShafique AhmedWei-Ho ChungYu TsaoHsin-Min WangPublished in: O-COCOSDA (2021)
Keyphrases
- single channel
- end to end
- frequency domain
- speech enhancement
- noisy environments
- automatic speech recognition
- speech signal
- spatial domain
- multi channel
- speech recognition
- denoising
- hidden markov models
- sound source
- noise reduction
- feature extraction
- subband
- wavelet domain
- wiener filter
- vocal tract
- signal to noise ratio
- prior information
- independent component analysis
- signal processing
- language model
- brain activity
- high quality
- image sequences