AI Voice Translator with Emotion Recognition

Introduction

Communication is more than just words—it’s tone and emotion. For my final year project, I built an AI-based voice translator that detects emotion and tone during real-time translation.

Core Features

• Speech-to-text translation.

• Emotion detection using deep learning models (TensorFlow & PyTorch).

• Personalized voice synthesis to replicate the speaker’s tone.

• Low-latency processing for live interactions.

Architecture

• Frontend: React.js for UI.

• Backend: Node.js for API handling.

• AI: Python + TensorFlow for ASR (Automatic Speech Recognition) and emotion models.

• APIs: Google Translate & Text-to-Speech APIs.

Impact

The project showcased how AI can create human-like, emotional communication across languages, and it was optimized for visually impaired users.