The Development and Integration of the LDA-Toolkit Into COST249 SpeechDat(II) SIG Reference Recognizer
نویسندگان
چکیده
This paper presents the development of Linear Discriminant Analysis toolkit (LDA-Toolkit) and its integration into widely used COST249 SpeechDat(II) Task Force Reference Recognizer (RefRec). The crucial parts of the LDA, the determination of LDA classes, as well as the influence of the level of dimensionality reduction on automatic speech recognition performance, are discussed. Evaluation of proposed LDA-RefRec procedure is performed using the Slovenian, German, and Spanish SpeechDat (II) databases. HTK (Hidden Markov Model Toolkit) is used in training and recognition processes. Features are computed using Advanced Front End (AFE) feature extraction procedure, proposed by Motorola, France Telecom, and Alcatel (AFE has been also standardized by ETSI organization). Automatic speech recognition results achieved with LDA-RefRec procedure show performance improvement and simultaneously dimensionality reduction when compared to baseline RefRec procedure. Proposed multilingual LDA classes, equal for all the three databases, perform only slightly worse than monolingual LDA classes, constructed and used separately for particular database. The results show benefits of the usage of the proposed LDA-RefRec procedure for evaluation or development of the automatic speech recognition systems based on SpeechDat (II) compliant databases.
منابع مشابه
An Isolated Letter Recognizer for Proper Name Identification Over the Telephone
Spelled letter recognition over the telephone line is essential for applications that involve names or addresses. In this paper we discuss the implementation and present results of a speaker independent spelled letter recognizer, trained and tested on the European project SPEECHDAT corpus. The system was implemented using HTK V2.0 (Hidden Markov Model Toolkit) software development tool and the ...
متن کاملThe COST 249 SpeechDat Multilingual Reference Recogniser
The COST 249 SpeechDat reference recogniser is a fully automatic, language-independent training procedure for building a phonetic recogniser. It relies on the HTK toolkit and a SpeechDat(II) compatible database. The recogniser is designed to serve as a reference system in multilingual recognition research. This paper documents version 0.93 of the reference recogniser and presents results on sma...
متن کاملFree Acoustic and Language Models for Large Vocabulary Continuous Speech Recognition in Swedish
This paper presents results for large vocabulary continuous speech recognition (LVCSR) in Swedish. We trained acoustic models on the public domain NST Swedish corpus and made them freely available to the community. The training procedure corresponds to the reference recogniser (RefRec) developed for the SpeechDat databases during the COST249 action. We describe the modifications we made to the ...
متن کاملA Noise Robust Multilingual Reference Recogniser Based on Speechdat(II)
An important aspect of noise robustness of automatic speech recognisers (ASR) is the proper handling of non-speech acoustic events. The present paper describes further improvements of an already existing reference recogniser towards achieving such kind of robustness. The reference recogniser applied is the COST 249 SpeechDat reference recogniser, which is a fully automatic, language-independent...
متن کاملHMM/MLP Hybrid Speech Recognizer for the Portuguese Telephone SpeechDat Corpus
In this article, we describe an automatic speech recognizer developed for Portuguese telephone speech. For this, we employed the Portuguese SpeechDat database which will be described in detail, giving its recording conditions, speaker characteristics and contents categories. The automatic recognizer is a state-of-the-art HMM/MLP hybrid system employing different kinds of robust acoustic feature...
متن کامل