Handwriting Generation and Synthesis using Recurrent Neural Networks

Handwritten text has been a vital mode of communication and a cornerstone of our culture and education for decades, and it is frequently regarded as an art form. It’s been found to help in tasks like taking notes and reading while writing, as well as improving short and long-term memory. The goal of handwriting synthesis is to create artificial text that closely resembles the user’s writing style. Handwriting synthesis not only adds a personal touch or preserves a user’s style, but it also has a number of other applications, such as improving text recognition systems, personalizing fonts, identifying writers, and spreading as the technology gets more popular. This paper presents a deep learning model for synthesizing handwriting text from computer text input. The goal is to create an AI tool that can generate handwriting while keeping a person’s style, which could potentially help people suffering with Dysgraphia. In addition, this paper compares our model to Alex Graves’ existing handwriting model.