Spectral-temporal receptive fields and MFCC balanced feature extraction for robust speaker recognition.
Jia-Ching WangChien-Yao WangYu-Hao ChinYu-Ting LiuEn-Ting ChenPao-Chi ChangPublished in: Multim. Tools Appl. (2017)
Keyphrases
- speaker recognition
- receptive fields
- feature extraction
- mel frequency cepstral coefficients
- speaker identification
- gaussian mixture model
- speaker verification
- spectral features
- biologically inspired
- vector quantization
- noisy environments
- probabilistic neural network
- speech signal
- speech recognition
- natural images
- space time
- feature vectors
- feature space
- image classification
- feature set
- pattern recognition
- pattern classification
- principal component analysis
- spatio temporal
- face recognition
- feature selection
- neural network
- multi modal
- broadcast news