Technical Program

Paper Detail

Paper:SP-P1.2
Session:Acoustic Modeling: Discriminative methods
Location:Poster Area D
Session Time:Tuesday, May 28, 10:50 - 12:50
Presentation Time:Tuesday, May 28, 10:50 - 12:50
Presentation: Poster
Topic: Speech Processing: Acoustic Modeling for Automatic Speech Recognition (SPE-RECO)
Paper Title: TIED-STATE BASED DISCRIMINATIVE TRAINING OF CONTEXT-EXPANDED REGION-DEPENDENT FEATURE TRANSFORMS FOR LVCSR
Authors: Zhi-Jie Yan, Qiang Huo, Jian Xu, Yu Zhang, Microsoft Research Asia, China