selected publications blog posting Self-attention encoding and pooling for speaker recognition. arXiv (Cornell University). 2020