Development of real-time embedded system with speech recognition for smart house
The hidden Markov model is used for the acoustic modeling of the speech recognition system. The Continuous Density HMM (CDHMM) which models acoustic observation directly using estimated continuous probability density function (pdf) without VQ, has shown to have higher recognition accuracy than DHMM....
| Main Authors: | , , |
|---|---|
| Format: | Monograph |
| Language: | English |
| Published: |
Faculty of Electric Engieering
2007
|
| Subjects: | |
| Online Access: | http://eprints.utm.my/6700/ http://eprints.utm.my/6700/1/74257.pdf |
| _version_ | 1848891328993165312 |
|---|---|
| author | Yahya, Zuraimi Sheikh Salleh, Sheikh Hussain Syed Abu Bakar, Syed Abdul Rahman |
| author_facet | Yahya, Zuraimi Sheikh Salleh, Sheikh Hussain Syed Abu Bakar, Syed Abdul Rahman |
| author_sort | Yahya, Zuraimi |
| building | UTeM Institutional Repository |
| collection | Online Access |
| description | The hidden Markov model is used for the acoustic modeling of the speech recognition system. The Continuous Density HMM (CDHMM) which models acoustic observation directly using estimated continuous probability density function (pdf) without VQ, has shown to have higher recognition accuracy than DHMM. In this paper, CDHMM with Multivariate Gaussian Density is used for speaker dependent (SD) and speaker independent (SI) Malay isolated digit recognition and comparison is made with DHMM. The CDHMM was trained by different algorithms- Baum-Welch (BW), Viterbi (VB) (with segmental k-mean estimation) algorithm and combination of BW and VB then comparison is discussed. The training database consisted of 26 speakers with 5 utterances for each Malay digit. In SD task, another 5 utterances each digit from the same speakers are used for testing. Recognition accuracy of CDHMM with BW, VB training and combination of BW and VB is 99.00%, 98.85% and 98.69% respectively while 96.62% for DHMM. The accuracy for BW and Segmental K-Mean training is comparable, but the latter consumed less computational time. In SI task, 40 speakers, different from the training speakers, with each recorded 2 tokens are used for testing. The CDHMM achieves 85.63% accuracy and outperform DHMM with 8.17% improvement. |
| first_indexed | 2025-11-15T20:56:14Z |
| format | Monograph |
| id | utm-6700 |
| institution | Universiti Teknologi Malaysia |
| institution_category | Local University |
| language | English |
| last_indexed | 2025-11-15T20:56:14Z |
| publishDate | 2007 |
| publisher | Faculty of Electric Engieering |
| recordtype | eprints |
| repository_type | Digital Repository |
| spelling | utm-67002017-07-25T02:18:25Z http://eprints.utm.my/6700/ Development of real-time embedded system with speech recognition for smart house Yahya, Zuraimi Sheikh Salleh, Sheikh Hussain Syed Abu Bakar, Syed Abdul Rahman TK Electrical engineering. Electronics Nuclear engineering The hidden Markov model is used for the acoustic modeling of the speech recognition system. The Continuous Density HMM (CDHMM) which models acoustic observation directly using estimated continuous probability density function (pdf) without VQ, has shown to have higher recognition accuracy than DHMM. In this paper, CDHMM with Multivariate Gaussian Density is used for speaker dependent (SD) and speaker independent (SI) Malay isolated digit recognition and comparison is made with DHMM. The CDHMM was trained by different algorithms- Baum-Welch (BW), Viterbi (VB) (with segmental k-mean estimation) algorithm and combination of BW and VB then comparison is discussed. The training database consisted of 26 speakers with 5 utterances for each Malay digit. In SD task, another 5 utterances each digit from the same speakers are used for testing. Recognition accuracy of CDHMM with BW, VB training and combination of BW and VB is 99.00%, 98.85% and 98.69% respectively while 96.62% for DHMM. The accuracy for BW and Segmental K-Mean training is comparable, but the latter consumed less computational time. In SI task, 40 speakers, different from the training speakers, with each recorded 2 tokens are used for testing. The CDHMM achieves 85.63% accuracy and outperform DHMM with 8.17% improvement. Faculty of Electric Engieering 2007-05-10 Monograph NonPeerReviewed application/pdf en http://eprints.utm.my/6700/1/74257.pdf Yahya, Zuraimi and Sheikh Salleh, Sheikh Hussain and Syed Abu Bakar, Syed Abdul Rahman (2007) Development of real-time embedded system with speech recognition for smart house. Project Report. Faculty of Electric Engieering, Skudai, Johor. (Unpublished) http://www.penerbit.utm.my |
| spellingShingle | TK Electrical engineering. Electronics Nuclear engineering Yahya, Zuraimi Sheikh Salleh, Sheikh Hussain Syed Abu Bakar, Syed Abdul Rahman Development of real-time embedded system with speech recognition for smart house |
| title | Development of real-time embedded system with speech recognition for smart house |
| title_full | Development of real-time embedded system with speech recognition for smart house |
| title_fullStr | Development of real-time embedded system with speech recognition for smart house |
| title_full_unstemmed | Development of real-time embedded system with speech recognition for smart house |
| title_short | Development of real-time embedded system with speech recognition for smart house |
| title_sort | development of real-time embedded system with speech recognition for smart house |
| topic | TK Electrical engineering. Electronics Nuclear engineering |
| url | http://eprints.utm.my/6700/ http://eprints.utm.my/6700/ http://eprints.utm.my/6700/1/74257.pdf |