Label-free Knowledge Distillation with Contrastive Loss for Light-weight Speaker Recognition.

Zhiyuan Peng Xuanji He Ke Ding Tan Lee Guanglu Wan

Published in: CoRR (2022)

Keyphrases

lightweight
speaker recognition
prior knowledge
vector quantization
gaussian mixture model
machine learning
information retrieval
image processing
pattern recognition
feature vectors
edge detection
signal to noise ratio
speaker identification
probabilistic neural network