| Paper: | SP-P14.2 | ||
| Session: | Acoustic Modeling: Tone, Prosody, and Features | ||
| Time: | Thursday, May 20, 15:30 - 17:30 | ||
| Presentation: | Poster | ||
| Topic: | Speech Processing: Acoustic Modeling for Speech Recognition | ||
| Title: | DISCRIMINATIVE FEATURE TRANSFORMATION BY GUIDED DISCRIMINATIVE TRAINING | ||
| Authors: | Roger Hsiao; Hong Kong University of Science and Technology | ||
| Brian Mak; Hong Kong University of Science and Technology | |||
| Abstract: | In this paper, we investigate a special form of discriminative training which we call guided discriminative training in the context of multi-class classification problems. We are interested in applications that require improving the classification performance of only a subset of the classes at the expense of possibly (but not necessarily) poorer classification performance of the remaining classes. However, should the classification of the remaining classes get worse, it is guaranteed not to be worse than an extent that the user specifies. The problem is formulated as a nonlinear programming problem, which can be translated to a unconstrained nonlinear optimization problem using the barrier method that, in turn, can be solved by gradient descent method. To prove the concept, we applied the guided discriminative training to derive an optimal linear transformation on the mel-filterbank log power spectra to improve TIMIT phoneme classification. Encouraging results are obtained. | ||
| Back | |||
Home -||-
Organizing Committee -||-
Technical Committee -||-
Technical Program -||-
Plenaries
Paper Submission -||-
Special Sessions -||-
ITT -||-
Paper Review -||-
Exhibits -||-
Tutorials
Information -||-
Registration -||-
Travel Insurance -||-
Housing -||-
Workshops