This project implements a multi-class classification model to classify Indian food items using Convolutional Neural Networks (CNN). The model is built with PyTorch and trained on a dataset of Indian food images.
The goal of this project is to build a model that can classify various types of Indian food. The classification is performed using a Convolutional Neural Network (CNN) architecture, which is trained to recognize and categorize images of Indian food items.
- Dataset: A collection of Indian food images for multi-class classification.
- Framework: Built using PyTorch.
- Data Augmentation: Used trivial augmentations and random horizontal transformations to increase the robustness of the model.
- Model Architecture: A CNN model with approximately 1.3 million parameters.
- Optimizer: Stochastic Gradient Descent (SGD) optimizer for model training.
- Loss Function: CrossEntropyLoss for multi-class classification.
- Visualization: Used Matplotlib to plot and visualize sample images during training.
- Python
- PyTorch
- Matplotlib
- NumPy
- PIL (Python Imaging Library)