CNN Image Emotion Classification

This repository contains deep learning models for image emotion classification using Convolutional Neural Networks (CNN).

Contents

• Efficientnetb4.ipynb - EfficientNet B4 model implementation for emotion classification
• RESET18.ipynb - ResNet-18 model implementation for emotion classification

Overview

This project explores different CNN architectures for classifying emotions in images. The repository includes implementations of:

1. EfficientNet B4: A state-of-the-art CNN architecture known for its efficiency and performance
2. ResNet-18: A residual neural network with 18 layers, popular for image classification tasks

Usage

1. Clone this repository
2. Open the Jupyter notebooks in your preferred environment
3. Install required dependencies (listed in each notebook)
4. Run the cells to train and evaluate the models

Models

EfficientNet B4

• Pre-trained model fine-tuned for emotion classification
• Optimized for both accuracy and computational efficiency

ResNet-18

• Deep residual learning architecture
• Proven performance on image classification tasks

Requirements

• Python 3.x
• TensorFlow/Keras or PyTorch (depending on implementation)
• Jupyter Notebook
• NumPy, Pandas, Matplotlib
• Additional dependencies as specified in notebooks

License

This project is for educational purposes.