Acoustic Similarity Scores for Keyword Spotting

Autor:	Luis A. S. V. de Sa, Fernando Perdigão, Arlindo Veiga, Carla Teixeira Lopes
Rok vydání:	2014
Předmět:	Normalization (statistics) Computer science business.industry Keyword spotting Speech recognition InformationSystems_INFORMATIONSTORAGEANDRETRIEVAL Pattern recognition Artificial intelligence business Word length
Zdroj:	Lecture Notes in Computer Science ISBN: 9783319097602 PROPOR
DOI:	10.1007/978-3-319-09761-9_5
Popis:	This paper presents a study on keyword spotting systems based on acoustic similarity between a filler model and keyword model. The ratio between the keyword model likelihood and the generic (filler) model likelihood is used by the classifier to detect relevant peaks values that indicate keyword occurrences. We have changed the standard scheme of keyword spotting system to allow keyword detection in a single forward step. We propose a new log-likelihood ratio normalization to minimize the effect of word length on the classifier performance. Tests show the effectiveness of our normalization method against two other methods. Experiments were performed on continuous speech utterances of the Portuguese TECNOVOZ database (read sentences) with keywords of several lengths.
Databáze:	OpenAIRE
Externí odkaz:	https://explore.openaire.eu/search/publication?articleId=doi_________::737db89e73271638c6a6351f8a59b373 https://doi.org/10.1007/978-3-319-09761-9_5 Zobrazit plný text záznamu