Keyword spotting system for broadcast news
A two-step keyword spotting strategy was presented.This strategy allowed to change keyword list conven-iently.Different from the previous systems,search space was generated based on all Chinese syllables,not specifically for keywords.Phoneme recognition was performed without any lexical constraints....
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | zho |
| Published: |
Editorial Department of Journal on Communications
2007-01-01
|
| Series: | Tongxin xuebao |
| Online Access: | http://www.joconline.com.cn/thesisDetails?columnId=74658803&Fpath=home&index=0 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850213674813227008 |
|---|---|
| author | ZHANG Peng-yuan SHAO Jian ZHAO Qing-wei YAN Yong-hong |
| author_facet | ZHANG Peng-yuan SHAO Jian ZHAO Qing-wei YAN Yong-hong |
| author_sort | ZHANG Peng-yuan |
| collection | DOAJ |
| description | A two-step keyword spotting strategy was presented.This strategy allowed to change keyword list conven-iently.Different from the previous systems,search space was generated based on all Chinese syllables,not specifically for keywords.Phoneme recognition was performed without any lexical constraints.With 1-best phoneme sequence and keyword list,which generated keyword hypotheses.At last,two confidence measures were introduced adopted in the sys-tem: one based on acoustic model and the other based on phoneme lattice.For a decoded speech frame aligned to an HMM state,the acoustic confidence was calculated.The lattice confidence made use of phoneme lattices generated by a phoneme recognizer.These two confidence measures were combined using a weighting factor to obtain a hybrid confi-dence as they had different dynamic scales.Experiments show that the proposed algorithms significantly improve the system performance on the broadcast news task. |
| format | Article |
| id | doaj-art-d0aa15bf34f44db4945fd2d1b79f0a43 |
| institution | OA Journals |
| issn | 1000-436X |
| language | zho |
| publishDate | 2007-01-01 |
| publisher | Editorial Department of Journal on Communications |
| record_format | Article |
| series | Tongxin xuebao |
| spelling | doaj-art-d0aa15bf34f44db4945fd2d1b79f0a432025-08-20T02:09:05ZzhoEditorial Department of Journal on CommunicationsTongxin xuebao1000-436X2007-01-0113113574658803Keyword spotting system for broadcast newsZHANG Peng-yuanSHAO JianZHAO Qing-weiYAN Yong-hongA two-step keyword spotting strategy was presented.This strategy allowed to change keyword list conven-iently.Different from the previous systems,search space was generated based on all Chinese syllables,not specifically for keywords.Phoneme recognition was performed without any lexical constraints.With 1-best phoneme sequence and keyword list,which generated keyword hypotheses.At last,two confidence measures were introduced adopted in the sys-tem: one based on acoustic model and the other based on phoneme lattice.For a decoded speech frame aligned to an HMM state,the acoustic confidence was calculated.The lattice confidence made use of phoneme lattices generated by a phoneme recognizer.These two confidence measures were combined using a weighting factor to obtain a hybrid confi-dence as they had different dynamic scales.Experiments show that the proposed algorithms significantly improve the system performance on the broadcast news task.http://www.joconline.com.cn/thesisDetails?columnId=74658803&Fpath=home&index=0 |
| spellingShingle | ZHANG Peng-yuan SHAO Jian ZHAO Qing-wei YAN Yong-hong Keyword spotting system for broadcast news Tongxin xuebao |
| title | Keyword spotting system for broadcast news |
| title_full | Keyword spotting system for broadcast news |
| title_fullStr | Keyword spotting system for broadcast news |
| title_full_unstemmed | Keyword spotting system for broadcast news |
| title_short | Keyword spotting system for broadcast news |
| title_sort | keyword spotting system for broadcast news |
| url | http://www.joconline.com.cn/thesisDetails?columnId=74658803&Fpath=home&index=0 |
| work_keys_str_mv | AT zhangpengyuan keywordspottingsystemforbroadcastnews AT shaojian keywordspottingsystemforbroadcastnews AT zhaoqingwei keywordspottingsystemforbroadcastnews AT yanyonghong keywordspottingsystemforbroadcastnews |