Improve recognition performance of limabeam algorithm using mllr adaptation for environment
This paper presents method using MLLR adaptation to improve recognition performance of Limabeam algorithm in speech recognition using microphone array for Korean database. From our investigation for this Limabeam, we could see that because the performance of filtering optimization depends strongly o...
Gardado en:
| Những tác giả chính: | , |
|---|---|
| Formato: | Bài viết |
| Idioma: | English |
| Publicado: |
Trường Đại học Đà Lạt
2012
|
| Những chủ đề: | |
| Acceso en liña: | https://scholar.dlu.edu.vn/thuvienso/handle/DLU123456789/33633 |
| Các nhãn: |
Engadir etiqueta
Sen Etiquetas, Sexa o primeiro en etiquetar este rexistro!
|
| Thư viện lưu trữ: | Thư viện Trường Đại học Đà Lạt |
|---|
| Tóm tắt: | This paper presents method using MLLR adaptation to improve recognition performance of Limabeam algorithm in speech recognition using microphone array for Korean database. From our investigation for this Limabeam, we could see that because the performance of filtering optimization depends strongly on the supporting optimal state sequence. This sequence was created by using Viterbi algorithm with HMM trained model. So the proposed approach is based on that we used MLLR adaptation for the utterance in new environment to obtain the better optimal state sequence that support for the filtering parameters optimal step.
Experimental results showed that using MLLR adaptation embed into the system, we have got the word correct recognition rate 2% higher than that of original calibrate Limabeam also 7% higher compared to Delay and Sum algorithm. |
|---|