Audio processing on constrained devices

Speech processing;Audio processing on mobile phones;Mobile phone applications;Computer Science

Gupta, Amod

University of Waterloo

关键词: Speech processing;    Audio processing on mobile phones;    Mobile phone applications;    Computer Science;   

Others  :  https://uwspace.uwaterloo.ca/bitstream/10012/4830/1/Thesis.pdf

瑞士|英语

来源: UWSPACE Waterloo Institutional Repository

【 摘 要 】

This thesis discusses the future of smart business applications on mobile phonesand the integration of voice interface across several business applications. It proposesa framework that provides speech processing support for business applicationson mobile phones. The framework uses Gaussian Mixture Models (GMM)for low-enrollment speaker recognition and limited vocabulary speech recognition.Algorithms are presented for pre-processing of audio signals into different categoriesand for start and end point detection. A method is proposed for speech processingthat uses Mel Frequency Cepstral Coeffcients (MFCC) as primary feature for extraction.In addition, optimization schemes are developed to improve performance,and overcome constraints of a mobile phone. Experimental results are presentedfor some prototype applications that evaluate the performance of computationallyexpensive algorithms on constrained hardware. The thesis concludes by discussingthe scope for improvement for the work done in this thesis and future directions inwhich this work could possibly be extended.