| Paper: | SS-5.3 | ||
| Session: | Dealing with Intrinsic Speech Variabilities in ASR | ||
| Time: | Wednesday, May 17, 14:40 - 15:00 | ||
| Presentation: | Special Session Lecture | ||
| Topic: | Special Sessions: Dealing with intrinsic speech variabilities in ASR | ||
| Title: | Characterizing Feature Variability in Automatic Speech Recognition Systems | ||
| Authors: | Loïc Barrault, Driss Matrouf, Renato De Mori, University of Avignon, France; Roberto Gemello, Franco Mana, Loquendo, Italy | ||
| Abstract: | A method is described for predicting acoustic feature variability by analyzing the consensus and relative entropy of phoneme posterior probability distributions obtained with different acoustic models having the same type of observations. Variability prediction is used for diagnosis of automatic speech recognition (ASR) systems. When errors are likely to occur, different feature sets are considered for correcting recognition results. Experimental results are provided on the CH1 Italian portion of AURORA3. | ||
©2018 Conference Management Services, Inc. -||- email: webmaster@icassp2006.org -||- Last updated Friday, August 17, 2012