Skip to content

Latest commit

 

History

History
45 lines (23 loc) · 1.83 KB

README.md

File metadata and controls

45 lines (23 loc) · 1.83 KB
Raw

Image Classification Project

Overview

This project focuses on initiating a machine learning project for image classification, specifically comparing the performance of the Vision Transformer and a Convolutional Neural Network (CNN). The experiment is conducted using Google Colab as the environment, and the CIFAR-100 image dataset is employed to train both models. The models are implemented using Keras and TensorFlow in Python.

Objectives

  • Model Comparison: Evaluate and compare the performance of the Vision Transformer and CNN in image classification.

  • Dataset: Utilize the CIFAR-100 image dataset for training both models.

  • Tools: Implement the models using Keras and TensorFlow in Python.

Experiment Results

The experiment revealed that, with a dataset of 60,000 images, the CNN demonstrated slightly better performance than the Vision Transformer overall.

Project Structure

The project includes the following components:

  • Code Snippet: The provided code initializes the machine learning project for image classification.

  • Models: The Vision Transformer and CNN are built using Keras and TensorFlow.

  • Dataset: CIFAR-100 image dataset is used for training.

How to Use

  1. Environment Setup: Use Google Colab as the development environment.

  2. Dataset Download: Acquire the CIFAR-100 dataset for training.

  3. Code Implementation: Explore and run the code snippet to initialize the project.

  4. Model Training: Train both the Vision Transformer and CNN models separately.

  5. Performance Comparison: Analyze the training results to compare the performance of the two models.

Conclusion

In summary, this project provides insights into the comparative performance of Vision Transformer and CNN in image classification, using the CIFAR-100 dataset.

Happy Experimenting! 🚀🖼️🤖