CASA Based Speech Separation for Robust Speech Recognition

DMCA

CASA Based Speech Separation for Robust Speech Recognition

Cached

Download Links

by Han Runqiang , Zhao Pei , Gao Qin , Zhang Zhiping , Wu Hao , Wu Xihong

Citations:

2 - 0 self

BibTeX

@MISC{Runqiang_casabased,
    author = {Han Runqiang and Zhao Pei and Gao Qin and Zhang Zhiping and Wu Hao and Wu Xihong},
    title = {CASA Based Speech Separation for Robust Speech Recognition},
    year = {}
}

OpenURL

Abstract

This paper introduces a speech separation system as a front-end processing step for automatic speech recognition (ASR). It employs computational auditory scene analysis (CASA) to separate the target speech from the interference speech. Specifically, the mixed speech is preprocessed based on auditory peripheral model. Then a pitch tracking is conducted and the dominant pitch is used as a main cue to find the target speech. Next, the time frequency (TF) units are merged into many segments. These segments are then combined into streams via CASA initial grouping. A regrouping strategy is employed to refine these streams via amplitude modulate (AM) cues, which are finally organized by the speaker recognition techniques into corresponding speakers. Finally, the output streams are reconstructed to compensate the missing data in the abovementioned processing steps by a cluster based feature reconstruction. Experimental results of ASR show that at low TMR (<-6dB) the proposed method offers significantly higher recognition accuracy. Index term: speech separation, CASA, speaker recognition, pitch tracking, units grouping, reconstruction 1.

Keyphrases

speech separation robust speech recognition pitch tracking target speech auditory peripheral model low tmr abovementioned processing step regrouping strategy time frequency automatic speech recognition main cue feature reconstruction front-end processing step dominant pitch many segment index term mixed speech speaker recognition computational auditory scene analysis output stream interference speech amplitude modulate recognition accuracy speaker recognition technique casa initial grouping speech separation system experimental result

DMCA