TY  - RPRT
TI  - Improving Unsupervised Sparsespeech Acoustic Models with Categorical Reparameterization
AU  - Benjamin Milde
AU  - Chris Biemann
PY  - 2020
UR  - https://arxiv.org/abs/2005.14578
ID  - 2005.14578
ER  -