A Hybrid HMM-Based Speech Recognizer Using Kernel-Based Discriminants as Acoustic Models

In this paper we propose a novel order-recursive training algorithm for kernel-based discriminants which is computationally efficient. We integrate this method in a hybrid HMM-based speech recognition system by translating the outputs of the kernel-based classifier into class-conditional probabilities and using them instead of Gaussian mixtures as production probabilities of a HMM-based decoder for speech recognition. The performance of the described hybrid structure is demonstrated on the DARPA Resource Management (RM1) corpus.

• Type: Conference
• Year: 2006
• Where: ICPR
• Authors: Edin Andelic, Marcel Katz, Martin Schafföner, Sven E. Krüger
• Keywords: Computer Vision, HMM-based Decoder, Hybrid Hmm-based Speech, ICPR 2006, Speech Recognition