🧠 CNN for Image Classification (MNIST & Fashion-MNIST) 📌 Project Overview
This project implements a Convolutional Neural Network (CNN) for image classification tasks using the MNIST and Fashion-MNIST datasets. MNIST: 28×28 grayscale images of handwritten digits (0–9). Fashion-MNIST: 28×28 grayscale images of fashion items (shirts, shoes, bags, etc.) across 10 categories. The CNN model is designed to automatically learn and recognize patterns from raw pixel data, making it effective for identifying digits and fashion objects.
Team Information:
| Name | Institute | |
|---|---|---|
| Niraj Band | GHRCE | niraj.band.it.@ghrce.raisoni.net |
| Tejaswini Badpaiya | GHRCE | tejaswini.badpaiya.it.@ghrce.raisoni.net |
| Swikruti Thantharate | GHRCE | swikruti.thantharate.ds@ghrce.raisoni.net |
| Riya Darda | GHRCE | riya.darda.it.@ghrce.raisoni.net |
| Sakshi Kadu | GHRCE | sakshi.kadu.it.@ghrce.raisoni.net |
| Sania Sheikh | GHRCE | sania.sheikh.it.@ghrce.raisoni.net |
🧑🏫 Mentor
Mr. Anirvinya Gururajan International Institute of Information Technology, Hyderabad, Telangana, India
⚙️ Features
Deep learning model using Convolutional Neural Networks (CNNs) Supports MNIST and Fashion-MNIST datasets High accuracy in recognizing handwritten digits and fashion items Easy to extend for other image classification tasks
📊 Datasets
MNIST: 60,000 training images + 10,000 testing images of digits (0–9). Fashion-MNIST: 60,000 training images + 10,000 testing images across 10 clothing categories. Both datasets are available in TensorFlow and PyTorch libraries.
📈 Results (Expected)
MNIST: >98% accuracy Fashion-MNIST: ~90% accuracy
✨ Made with passion by the GHRCE Team under the guidance of Mr. Anirvinya Gururajan