Discovering Linguistic Structures in Speech: Models and Applications

Jackie Lee

Kite

Friday, November 6, 2015
11:00 a.m., ICSI Lecture Hall

The ability to infer linguistic structures from noisy speech streams seems to be an innate human capability. However, reproducing the same ability in machines has remained a challenging task. In this talk, I will show a class of probabilistic models that my collaborators and I developed for discovering latent linguistic structures of a language directly from acoustic signals. In particular, we explore a nonparametric Bayesian framework for automatically acquiring a phone-like inventory of a language. Furthermore, we integrate this phone discovery model with adaptor grammars, a nonparametric Bayesian extension of probabilistic context-free grammars, to induce hierarchical linguistic structures, including sub-word and word-like units, directly from speech signals. When tested on a variety of speech corpora containing different acoustic conditions, domains, and languages, these models consistently demonstrate an ability to learn highly meaningful linguistic structures.

Bio:

Jackie Lee received her MS and PhD degrees from MIT in 2010 and 2014, where she was a graduate student at the Spoken Language Systems group at CSAIL. Her research interests lie in the intersection of speech recognition, natural language processing and cognitive science. In particular, she enjoyed building unsupervised models to discover linguistic structures directly from speech data. Jackie is currently a machine learning engineer at Kite.