NLP-Based Text-to-speech and Speech-to-text Virtual Assistant

Artificial Intelligence (AI) technologies played a vital role in one's life. Analyzing human language by virtual assistants is one of the trends in AI. In this proposed work, Assistants recognize human language with the help of Speech Recognition and Natural Language Processing (NLP). This virtual assistant performs the task as per the commands given by the user in various domains like sending emails, watching videos on YouTube, writing in a word file/notepad, clicking pictures in the camera, getting the PDF file in an audiobook, etc. Moreover, to analyze the performance of this Assistant, an audio sample of 1000 sentences is given to the system, and analysis is done based on some parameters like pitch, voice, language, background noise, etc. to check its performance. The experimental results show that the proposed model achieves an accuracy of 97.7%.