Statistical language models estimate the probability of a word occurring in a given context. The most common language models rely on a discrete enumeration of predictive contexts ...
John Blitzer, Kilian Q. Weinberger, Lawrence K. Sa...
Learning data representations is a fundamental challenge in modeling neural processes and plays an important role in applications such as object recognition. In multi-stage Optima...
Abstract--Motivated by potential applications in second-language pedagogy, we present a novel approach to using articulatory information to improve automatic detection of typical p...