M-vectors: Sub-band Based Energy Modulation Features for Multi-stream Automatic Speech Recognition.
Samik SadhuRuizhi LiHynek HermanskyPublished in: ICASSP (2019)
Keyphrases
- feature vectors
- automatic speech recognition
- subband
- multi stream
- speech recognition
- feature extraction
- hidden markov models
- feature set
- wavelet transform
- discrete wavelet transform
- high frequency
- wavelet coefficients
- image processing
- low frequency
- frequency domain
- speech signal
- noisy environments
- broadcast news
- audio visual speech recognition
- bit rate
- multiresolution
- feature space
- information fusion
- filter bank
- image compression
- image features
- pattern recognition
- image segmentation