Design and implementation of smart voice assistant and recognizing academic words

This paper approaches the use of a Virtual Assistant using neural networks for recognition commonly used words. The main purpose is to facilitate users’ daily lives by sensing voice and interpreting it into action. Alice, which name assistant, implemented based on four techniques: Hot word detection, Voice Text conversion, Intent recognition, conversion. Linux operating system choice, developing running assistant because in public domain, also, has been most Single-board computers. Python chosen as development language due its capabilities compatibility with various APIs libraries, are deemed necessary project. virtual will be required communicate IoT devices. In addition, speech created order recognize significant technical An artificial network (ANN) different structure training algorithms utilized conjunction Mel Frequency Cepstral Coefficient (MFCC) feature extraction technique increase identification rate effectively find optimal performance. For purposes, Levenberg-Marquardt (LM) BGFS Quasi-Newton Resilient Backpropagation compared 10 MFCC, utilizing from 50 neurons increasing increments similarly 13MFCC done between neurons.