Improvement of Telephone Keyword Spotting Performance Using Linear Programming-Based Score Normalization

,; ,; ,; ,

Volume 7, Issue 2 (3-2011) JSDP 2011, 7(2): 37-48 | Back to browse issues page

Mendeley

Zotero

RefWorks

Improvement of Telephone Keyword Spotting Performance Using Linear Programming-Based Score Normalization. JSDP 2011; 7 (2) :37-48
URL: http://jsdp.rcisp.ac.ir/article-1-716-en.html

Improvement of Telephone Keyword Spotting Performance Using Linear Programming-Based Score Normalization

Abstract: (4471 Views)

Conventional word spotting systems determine hypothesized keywords and their confidence score using a speech recognizer. Acceptance or rejection of these keywords is intended based on comparison of their scores with a specific threshold. It has been proved that confidence score prepared by recognizer is highly dependent on sub-word structure of each keyword. So comparing assigned scores to keywords without considering their sub-word units could causes degradation in overall performance. In this paper a novel method for confidence score normalization is proposed which is based on sub-word units of each keyword and linear programming algorithm. In proposed method, a keyword-dependent correction term is added to the score of the keyword to maximize separation of confidence score histograms of true and false occurrences. Our results show a 2% improvement in FOM compared to baseline system. Also, choosing an appropriate feature vector has been discussed in this paper.

Keywords: Keyword spotting, Hidden Markov Model, Confidence score, Linear programming, Score normalization.

Full-Text [PDF 5517 kb] (978 Downloads)

Type of Study: Research | Subject: Paper
Received: 2011/03/11 | Accepted: 2018/02/19 | Published: 2018/02/19 | ePublished: 2018/02/19

Rights and permissions
	This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.

Signal and Data Processing

Vote