Malay speech recognition for different ethnic speakers : An exploratory study

Autor: Rafizah Mohd Hanifa, Khalid Isa, Shamsul Mohamad
Rok vydání: 2017
Předmět:
Zdroj: 2017 IEEE Symposium on Computer Applications & Industrial Electronics (ISCAIE).
DOI: 10.1109/iscaie.2017.8074956
Popis: Speech technology has become an alternative medium of communication. It could help those on smart wheelchair to use their voice to ease their movements from one location to another. However, most of the speech technology that has been developed is based on the English language and limited for other languages. This project contributes to the literature by developing a program on speech technology for Malay speech recognition using Visual C# to determine the quality of the program in advancing the state-of-the-art. The developed program has been tested using the speech by 13 subjects from different ethnic backgrounds to evaluate its quality and limitations. Results indicate that the voice, including the volume, accent and word emphasis to be important elements in influencing the quality of recognition.
Databáze: OpenAIRE