iTalk: A Conversational Chatbot with Personalized Voice Modulation for Emotional Aid

Abstract

"The field of artificial intelligence has experienced notable progress, namely in the area of conversational interfaces. In light of the global need to enhance mental and emotional well-being, there arises a noteworthy prospect of utilising technology as a means of resolving this issue. The focus of this research revolves around the creation of a conversational chatbot that aims to deliver personalised emotional support through messages and personalized real voice feedback by applying the capabilities of Large Language Models (LLMs) and the ElevenLabs Voice Service APIs. This project aims to explore the potential of Natural Language Processing techniques, LLMs, and voice cloning in order to gain a deeper understanding of the connection between technology and human emotion. The above-mentioned technologies are combined together to construct this novel solution called “iTalk” which is a web-application. This will serve the problem of loneliness found within each person by providing the users a platform to create and interact with personalized personas. Through the application, the user will be able to create a persona of another individual by responding to some pre-defined questions. Also “iTalk” enables voice cloning allowing the users to replicate the voice of a real person by providing a set of voice data. This would enable users to engage in conversations with the synthesized voice, and simulate the interactions as if the real person were speaking."