Unsupervised Subword Modeling Using Autoregressive Pretraining and Cross-Lingual Phone-Aware Modeling

This study addresses unsupervised subword modeling, i.e.,<br/>learning feature representations that can distinguish subword<br/>units of a language. The proposed approach adopts a two-stage<br/>bottleneck feature (BNF) learning framework, consisting of autoregressive<br/>predictive coding (APC) as a front-end and a DNNBNF<br/>model as a back-end...