Robust speaker recognition based on filtering in autocorrelation domain and sub-band feature recombination.
Sungtak KimMiyoung JiHoirin KimPublished in: Pattern Recognit. Lett. (2010)
Keyphrases
- speaker recognition
- subband
- feature vectors
- gaussian mixture model
- vector quantization
- high pass
- image compression
- discrete wavelet transform
- multiresolution
- wavelet coefficients
- frequency domain
- wavelet transform
- speaker verification
- high frequency
- filter bank
- bit rate
- low pass
- probabilistic neural network
- neural network
- machine learning
- speech recognition
- natural images
- feature extraction
- speaker identification
- image processing