Signal separation for robust speech recognition based on phase difference information obtained in the frequency domain.
Chanwoo KimKshitiz KumarBhiksha RajRichard M. SternPublished in: INTERSPEECH (2009)
Keyphrases
- frequency domain
- speech recognition
- fourier transform
- noisy environments
- spatial domain
- phase difference
- cross correlation
- pattern recognition
- power spectrum
- language model
- denoising
- automatic speech recognition
- single channel
- signal processing
- image compression
- probabilistic model
- hidden markov models
- feature extraction
- computer vision