INTEGRATING COMPUTATIONAL AUDITORY SCENE ANALYSIS AND AUTOMATIC SPEECH RECOGNITION

دانشگاه آزاد اسلامی اصفهان(خوراسگان)

منو

رکورد قبلی

رکورد بعدی

" INTEGRATING COMPUTATIONAL AUDITORY SCENE ANALYSIS AND AUTOMATIC SPEECH RECOGNITION "

Document Type	:	Latin Dissertation
Language of Document	:	English
Record Number	:	150120
Doc. No	:	ET21912
Main Entry	:	Soundararajan Srinivasan, M.S.
Title Proper	:	INTEGRATING COMPUTATIONAL AUDITORY SCENE ANALYSIS AND AUTOMATIC SPEECH RECOGNITION
Note	:	This document is digital این مدرک بصورت الکترونیکی می باشد
Abstract	:	Speech perception studies indicate that robustness of human speech recognitionis primarily due to our ability to segregate a target sound source from other inter-fering sources. This perceptual process of auditory scene analysis (ASA) is of twotypes, primitive and schema-driven. Primitive ASA is an innate mechanism basedon bottom-up acoustic cues such as pitch, common onset, and spatial location of asound source. Schema-based ASA involves the top-down use of stored knowledge,such as speech patterns, to supplement primitive analysis. This dissertation inves-tigates several aspects of integrating computational ASA (CASA) and automaticspeech recognition (ASR). While bottom-up CASA are used as front-end for ASRto improve its robustness, ASR is used to provide top-down information to enhanceprimitive segregation.,..tested for theQ1 PC1 bus cardBoth these projects mere sofixare des elopment efforts tonards contributing to dlfferentaspects of Roboucs and lZ1echatronics projects m the Controls and Roboucs Group..
Subject	:	Electericl tess
	:	برق
electronic file name	:	TL45106.pdf
Title and statement of responsibility and	:	INTEGRATING COMPUTATIONAL AUDITORY SCENE ANALYSIS AND AUTOMATIC SPEECH RECOGNITION [Thesis]