Speech-To-Text and Text-To-Speech applications are essentially based on an effective separation of phonetic units, so the segmentation of uttered speech into phonetic units is a key processing task for successfully implementing speech recognition systems. Softcomputing methods demonstrate to be more effective than other methods due to the capability neural networks and fuzzy logic to be trained by expert. This work phonetic segmentation of uttered speech that separates vowels from consonants is based on a fuzzy logic inference engine tuned by an expert using speech features distribution. Only time-domain feature-extraction algorithms are applied to speech to extract features, so minimum computational cost was achieved. Fuzzy decision logic is used to infer about phonetic units separation point. A set of tests has been executed to demonstrate that this approach can be effective in separating phonetic units, while requiring minimal computing power and reducing system complexity.

Softcomputing approach to segmentation of speech in phonetic units / M. Malcangi. - 3:3(2009), pp. 41-48.

Softcomputing approach to segmentation of speech in phonetic units

M. Malcangi
Primo
2009

Abstract

Speech-To-Text and Text-To-Speech applications are essentially based on an effective separation of phonetic units, so the segmentation of uttered speech into phonetic units is a key processing task for successfully implementing speech recognition systems. Softcomputing methods demonstrate to be more effective than other methods due to the capability neural networks and fuzzy logic to be trained by expert. This work phonetic segmentation of uttered speech that separates vowels from consonants is based on a fuzzy logic inference engine tuned by an expert using speech features distribution. Only time-domain feature-extraction algorithms are applied to speech to extract features, so minimum computational cost was achieved. Fuzzy decision logic is used to infer about phonetic units separation point. A set of tests has been executed to demonstrate that this approach can be effective in separating phonetic units, while requiring minimal computing power and reducing system complexity.
Fuzzy decision logic ; pitch estimation ; speech energy ; speech segmentation ; speech analysis ; speech recognition ; speech synthesis ; zero-crossing rate
Settore INF/01 - Informatica
2009
http://www.universitypress.org.uk/journals/cc/19-397.pdf
Article (author)
File in questo prodotto:
Non ci sono file associati a questo prodotto.
Pubblicazioni consigliate

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/2434/154051
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus ND
  • ???jsp.display-item.citation.isi??? ND
social impact