A Methodolgy for Realistic Handwriting Generation Using Multilayer Perceptrons and LSTM

Abstract

Hand-writing is one of the mainly imperative cryptographic parameters and a fundamental sign of people’s characteristics such as their personality. Hand writing generation is one of the prominent fields of research in view of the identification of persons personality. In this research paper the research is carried out for generation of handwriting using Multi-Layer Perceptron and Long Short-term Memory.In this task the fundamental focus is to exhibit how Multi-Layer perceptron can be utilized to produce complex groupings with long-range structure, essentially by anticipating each information point in turn. The proposed methodology is unveiled for text (where the information are discrete) and web based handwriting (where the information are genuine esteemed). It is then reached out to handwriting synthesis by permitting the organization to condition its forecasts on a text succession. The subsequent framework can produce exceptionally practical cursive handwriting in a wide assortment of styles.