Emotion-Aware Text-to-Speech Synthesis for Enhanced Accessibility: Synthetic Data Generation and Automated TTS for the Visually Impaired
M. Yagnasri Priya, Sanjivani P. Shendre, Samrudhi S. Kamble, Swetha Sridhar, Suja Palaniswamy · 2025
For effective communication, the ability to perceive and interpret emotions is fundamental. For visually impaired individuals, who rely heavily on auditory cues for digital content consumption, the lack of emotional nuances in the traditional Text-to-Speech(TTS) hinders their engagement and comprehension. While the existing TTS technologies focus primarily on clarity and naturalness, they fail to capture the emotional depth embedded in text, leaving a significant gap in solutions for accessibility. To address this, the work proposes an Emotion-Aware TTS system that enhances the accessibility and emotional comprehension of written content. The system employs advanced emotion recognition techniques to analyze and classify the emotional tone of text, coupled with synthetic data generation to overcome the scarcity of labeled emotional datasets. Developed in Python, this system is designed to enhance digital accessibility in educational, professional, and personal communication settings.