VOICE INTELLIGENCE BASED WAKE WORD DETECTION OF REGIONAL DIALECTS

VOICE INTELLIGENCE BASED WAKE WORD DETECTION OF REGIONAL DIALECTS

Chaitra GP, Dr.Shylaja S.S

Abstract— Technology is considered as a contributor to economic growth only when it reaches rural/remote areas where it can be effectively applied by farmers for their agricultural needs. Analyzing the factors affecting the adoption of technology by farmers plays a crucial role in updating the technology. Many voice-based apps were developed in the agricultural sector, and in each case, farmers had to either type in the queries or they had to communicate with the device which had the standard speech to deliver the solution Even if the farmers manage to utilize mobile phones to access the information, they may lack this ability to comprehend when it is not in their native language. If Technology must reach the rural farmers, then the fact about their connection towards the voice-based apps in their own spoken language/Dialect cannot be left unnoticed. This paper presents the research work in developing the wake word detection system for 4 of the major dialects in Karnataka. The customized wake word system is designed using CNN, TensorFlow along with Keras.

Keywords— TensorFlow, Keras API, Mel Frequency Cepstral Coefficients, CNN, Dialect Identification, Deep Learning Techniques, Sequential Modelling, DTW.