Analysis of Malay Speech Recognition for Different Speaker Origins
This paper explores speech recognition performance for Malay language with multi accents from speakers of different origins or ethnicities. Accented speech imposes accuracy problem in automatic speech recognition systems. This frequently occurs to non-native speakers of a language due to insufficien...
| Main Authors: | , , |
|---|---|
| Format: | Proceeding |
| Published: |
IEEE
2012
|
| Subjects: | |
| Online Access: | http://ir.unimas.my/id/eprint/8877/ |
| _version_ | 1848836462869479424 |
|---|---|
| author | Juan, Sarah Samson Besacier, Laurent Tan, Tien-Ping |
| author_facet | Juan, Sarah Samson Besacier, Laurent Tan, Tien-Ping |
| author_sort | Juan, Sarah Samson |
| building | UNIMAS Institutional Repository |
| collection | Online Access |
| description | This paper explores speech recognition performance for Malay language with multi accents from speakers of different origins or ethnicities. Accented speech imposes accuracy problem in automatic speech recognition systems. This frequently occurs to non-native speakers of a language due to insufficiency of the non-natives data in
the recognizers. In this study, we investigate the mentioned problem by building a Malay model in our recognizer and test its performance for speakers of various ethnicities. Our Malay corpora consist of read speeches and texts that are collected from local newspapers in Malaysia. Speakers who contributed the speeches are of different ethnic backgrounds. We employ context dependent models by applying linear discriminant analysis for our acoustic model and a trigram based language model. Our experiments show improved results when linear discriminant analysis technique was employed in our model while our recognizer performed worst for speakers with accent that are not available in the training data. |
| first_indexed | 2025-11-15T06:24:09Z |
| format | Proceeding |
| id | unimas-8877 |
| institution | Universiti Malaysia Sarawak |
| institution_category | Local University |
| last_indexed | 2025-11-15T06:24:09Z |
| publishDate | 2012 |
| publisher | IEEE |
| recordtype | eprints |
| repository_type | Digital Repository |
| spelling | unimas-88772015-10-16T01:19:03Z http://ir.unimas.my/id/eprint/8877/ Analysis of Malay Speech Recognition for Different Speaker Origins Juan, Sarah Samson Besacier, Laurent Tan, Tien-Ping T Technology (General) This paper explores speech recognition performance for Malay language with multi accents from speakers of different origins or ethnicities. Accented speech imposes accuracy problem in automatic speech recognition systems. This frequently occurs to non-native speakers of a language due to insufficiency of the non-natives data in the recognizers. In this study, we investigate the mentioned problem by building a Malay model in our recognizer and test its performance for speakers of various ethnicities. Our Malay corpora consist of read speeches and texts that are collected from local newspapers in Malaysia. Speakers who contributed the speeches are of different ethnic backgrounds. We employ context dependent models by applying linear discriminant analysis for our acoustic model and a trigram based language model. Our experiments show improved results when linear discriminant analysis technique was employed in our model while our recognizer performed worst for speakers with accent that are not available in the training data. IEEE 2012 Proceeding PeerReviewed Juan, Sarah Samson and Besacier, Laurent and Tan, Tien-Ping (2012) Analysis of Malay Speech Recognition for Different Speaker Origins. In: Proceedings of International Conference on Asian Language Processing (IALP). http://ieeexplore.ieee.org/xpl/articleDetails.jsp?arnumber=6473738 |
| spellingShingle | T Technology (General) Juan, Sarah Samson Besacier, Laurent Tan, Tien-Ping Analysis of Malay Speech Recognition for Different Speaker Origins |
| title | Analysis of Malay Speech Recognition for Different Speaker Origins |
| title_full | Analysis of Malay Speech Recognition for Different Speaker Origins |
| title_fullStr | Analysis of Malay Speech Recognition for Different Speaker Origins |
| title_full_unstemmed | Analysis of Malay Speech Recognition for Different Speaker Origins |
| title_short | Analysis of Malay Speech Recognition for Different Speaker Origins |
| title_sort | analysis of malay speech recognition for different speaker origins |
| topic | T Technology (General) |
| url | http://ir.unimas.my/id/eprint/8877/ http://ir.unimas.my/id/eprint/8877/ |