An Auditory Inspired Amplitude Modulation Filter Bank for Robust Feature Extraction in Automatic Speech Recognition.
Niko MoritzJörn AnemüllerBirger KollmeierPublished in: IEEE ACM Trans. Audio Speech Lang. Process. (2015)
Keyphrases
- filter bank
- automatic speech recognition
- feature extraction
- signal processing
- frequency domain
- subband
- speech recognition
- noisy environments
- image coding
- amplitude modulation
- computationally efficient
- perfect reconstruction
- multiscale
- wavelet transform
- multiresolution
- speech signal
- conversational speech
- image processing
- feature vectors
- hidden markov models
- face recognition
- feature space
- broadcast news
- pattern recognition
- information retrieval
- denoising
- visual information
- frequency modulation