TEXT TO SPEECH CONVERTER FOR HANDWRITTEN SCRIPT

Computer Engineering

Handwritten script recognition and conversion to speech has gained significant importance in various applications, including accessibility tools for the visually impaired, transcription services, and document automation. In this project, we present a Text-to-Speech (TTS) Converter for Handwritten Scripts, which seamlessly transforms handwritten text into spoken words. The system employs deep learning techniques, leveraging an Open Neural Network Exchange (ONNX)-based model for handwritten script recognition and Connectionist Temporal Classification (CTC) for efficient recognition of handwritten text sequences. The project's workflow initiates with the acquisition of handwritten scripts, which are subsequently processed through a trained ONNX model. The ONNX model, powered by convolutional neural networks (CNNs), effectively identifies individual characters within the handwritten text, even amidst varying styles and nuances. The usage of ONNX format facilitates platform-independent deployment, enabling the model to run efficiently across different environments. After successful recognition of the handwritten text, the project employs Google Text- to-Speech (gTTS) to convert the recognized text into natural and intelligible speech. The integration of gTTS ensures high-quality speech synthesis with support for multiple languages and voice options, enhancing the user experience. The Text-to-Speech Converter for Handwritten Scripts presented in this project represents a valuable tool for various domains, including education, assistive technology, and automation. Its accurate handwritten script recognition and efficient speech synthesis capabilities contribute to enhanced usability and accessibility for diverse user groups.

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.