Evaluating Acoustic Feature Maps in 2D-CNN for Speaker Identification.
Ali Shariq ImranVetle HaflanAbdolreza Sabzi ShahrebabakiNegar OlfatiTorbjørn Karl SvendsenPublished in: ICMLC (2019)
Keyphrases
- speaker identification
- feature maps
- audio signal
- speech recognition
- gaussian mixture model
- noisy environments
- feature vectors
- speech signal
- feature extraction
- self organizing maps
- kernel function
- broadcast news
- saliency map
- neural network
- feature space
- higher level
- multiscale
- mixture model
- training samples
- sound source
- training data