TY  - RPRT
TI  - A high speed unsupervised speaker retrieval using vector quantization and second-order statistics
AU  - Konstantin Biatov
PY  - 2010
UR  - https://arxiv.org/abs/1008.4658
ID  - 1008.4658
ER  -