CNN Research

This repository contains the codebase for my CNN architecture benchmarking research.

Overview

Trained and evaluated multiple convolutional neural networks — AlexNet, VGG16, ResNet50, and InceptionV3 — on the CIFAR-10 dataset to compare accuracy, efficiency, and model complexity.

Contents

• models/ – CNN architecture defs
• data_loader.py – Handles dataset import and preprocessing
• evaluate_models.py – Runs training and evaluation across architectures
• requirements.txt – Dependencies

Paper

https://zenodo.org/doi/10.5281/zenodo.13316277