https://scholars.lib.ntu.edu.tw/handle/123456789/121793
Title: | Improved robust features for speech recognition by integrating time-frequency principal components (TFPC) and histogram equalization (HEQ) | Authors: | Tsai, Shang-Nien LIN-SHAN LEE |
Issue Date: | Dec-2003 | Start page/Pages: | 297-302 | Source: | 2003 IEEE Workshop on Automatic Speech Recognition and Understanding, ASRU 2003 | Conference: | IEEE Workshop on Automatic Speech Recognition and Understanding, ASRU 2003 | Abstract: | Robustness for speech recognition technologies with respect to adverse environments has been a key issue for real applications. Time-frequency principal components (TFPC) features were shown to be a set of powerful data-driven features under matched circumstances, while histogram equalization (HEQ) was proposed as an efficient feature transformation approach to reduce the mismatch between training and testing conditions. In this paper, it is proposed that TFPC features can be well integrated with HEQ. HEQ generates a well-matched environment, in which TFPC features can be properly utilized. Extensive experiments with respect to the AURORA2 database verified that improved performance in adverse circumstances can be achieved. © 2003 IEEE. |
URI: | http://ntur.lib.ntu.edu.tw//handle/246246/200704191002819 http://ntur.lib.ntu.edu.tw/bitstream/246246/200704191002819/1/01318457.pdf https://www.scopus.com/inward/record.uri?eid=2-s2.0-33646523099&doi=10.1109%2fASRU.2003.1318457&partnerID=40&md5=58c16a312128bcb51ecf768efff8dfa8 |
Other Identifiers: | N/A | DOI: | 10.1109/ASRU.2003.1318457 | SDG/Keyword: | Graphic methods; Metadata; Adverse environment; Feature transformations; Histogram equalizations; Principal Components; Real applications; Speech recognition technology; Time frequency; Training and testing; Speech recognition |
Appears in Collections: | 電信工程學研究所 |
File | Description | Size | Format | |
---|---|---|---|---|
01318457.pdf | 465.15 kB | Adobe PDF | View/Open |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.