Decoding the Digits: Unleashing the Power of Convolutional Neural Networks for Handwritten Digit Recognition

Decoding the Digits: Unleashing the Power of Convolutional Neural Networks for Handwritten Digit Recognition

Aatmaj Amol Salunke

BACHELOR OF TECHNOLOGY

Computer Science & Engineering

Department of Computer Science & Engineering,

School of Computer Science and Engineering,

Manipal University Jaipur

Abstract

This research paper investigates the implementation of Convolutional Neural Networks (CNNs) for Handwritten Digit Recognition, with a specific emphasis on the popular MNIST dataset with a total of 70,000 samples. The abstract provides a succinct summary of our study. We thoroughly examine the CNN model's structure and training procedure, showcasing its efficacy in precisely categorizing handwritten digits. Through rigorous experimentation and assessment, we establish the model's superior accuracy and resilience. Our work contributes significantly to the domain of computer vision and opens up possibilities for leveraging CNNs in more sophisticated digit recognition tasks.

Keywords: Handwritten Digit Recognition, Convolutional Neural Networks (CNNs), MNIST dataset, Dense Layer, Accuracy, ReLU, Softmax.