Implementing a statistical method for automatic speech recognition
Date of Publication
1990
Document Type
Bachelor's Thesis
Degree Name
Bachelor of Science in Computer Science with Specialization in Computer Technology
College
College of Computer Studies
Department/Unit
Computer Technology
Abstract/Summary
Speech recognition centers on the use of natural speech for human-computer interaction providing computers an ear to listen to what human beings intend to say. In addition to speech recognition as being the most natural method of communication, it offers several advantages like ease of access, speed, manual freedom, and remote access. The Automatic Speech Recognition system is a prototype speaker-independent, isolated speech recognition system consisting of hardware and software components necessary in performance delivery. It was implemented using a statistical method that analyzes speech parameters to recognize sequence of words spoken by a user with pauses in-between words. Words uttered by the user are compared against the words trained and stored in the vocabulary file by computing likelihood probabilities based on speech characteristics extracted from the corresponding speech signals. The vocabulary word with the highest measure of likelihood is selected to be the most probable word uttered by the user. The accuracy of recognition depends primarily on the distinctiveness and the number of words in the vocabulary and the clarity with which the user says the words. The ASR as well as other speech recognition systems provide room for future applications. These applications include: (1) Clinical-Medical records, services for the handicapped (2) Entertainment and Education - Voice-controlled toys, interactive video games (3) Manufacturing Process Control - Machine operation, package sorting (4) Office Automation - Data entry, automatic dictation, automatic transcription and (5) Security - Voiceprint identification, building access.
Abstract Format
html
Language
English
Format
Accession Number
TU07970
Shelf Location
Archives, The Learning Commons, 12F, Henry Sy Sr. Hall
Physical Description
1 v. (various pagings)
Keywords
Automatic speech recognition; Systems software; Computer design; Speech processing systems
Recommended Citation
Gochuico, S., Lee, S., Marcos, N., & Yu, Y. (1990). Implementing a statistical method for automatic speech recognition. Retrieved from https://animorepository.dlsu.edu.ph/etd_bachelors/16405
