This paper takes a computational learning theory approach to a problem of linear systems identification. It is assumed that inputs are generated randomly from a known class consist...
Hidden Markov models hmms and partially observable Markov decision processes pomdps provide useful tools for modeling dynamical systems. They are particularly useful for represent...
Lexicalized reordering models play a crucial role in phrase-based translation systems. They are usually learned from the word-aligned bilingual corpus by examining the reordering ...
Abstract—We present an extension of automatic image annotation that takes the context of a picture into account. Our core assumption is that users do not only provide individual ...
The problem of acoustic-to-articulatory speech inversion continues to be a challenging research problem which significantly impacts automatic speech recognition robustness and ac...