🐾 Animal Image Classification System

📌 Project Overview

A deep learning-based multi-class image classification system built using PyTorch and EfficientNet-B0.

The model classifies images into 15 animal categories including: Bear, Bird, Cat, Cow, Deer, Dog, Dolphin, Elephant, Giraffe, Horse, Kangaroo, Lion, Panda, Tiger, Zebra.

All images are resized to 224x224x3 for training.

---

🎯 Objective

To build an accurate and scalable deep learning model capable of identifying animal species from images using transfer learning.

---

🧠 Model Architecture

• Transfer Learning using EfficientNet-B0
• Pretrained on ImageNet
• Custom classification head for 15 classes
• Fine-tuning after initial frozen training phase

---

🗂 Dataset

• 15 classes
• 224x224 RGB images
• Folder-based structure

Dataset is not included in this repository.

Place the dataset inside a folder named:

dataset/ ├── Bear/ ├── Bird/ ├── Cat/ ... └── Zebra/

Update the dataset path inside config.yaml if needed.

---

🛠 Tech Stack

• Python
• PyTorch
• Torchvision
• TIMM (for EfficientNet)
• NumPy
• OpenCV
• Streamlit (for deployment interface)

---

📊 Results

• Training Accuracy: (add your real number)
• Validation Accuracy: (add your real number)
• Model checkpoint saved as: models/best_model.pth

---

🚀 How to Run

1️⃣ Clone the repository: git clone https://github.com/shrashtimittal/animal-image-classification.git

cd animal-image-classification

2️⃣ Install dependencies: pip install -r requirements.txt

3️⃣ Install PyTorch separately (based on your system): Visit: https://pytorch.org/get-started/locally/

4️⃣ Run the application: streamlit run app.py

---

📌 Project Structure

animal-image-classification/ ├── src/ # Model architecture and training logic ├── models/ # Saved model weights ├── app.py # Streamlit app ├── config.yaml # Configuration file ├── requirements.txt └── README.md

---

🔮 Future Improvements

• Hyperparameter tuning
• Model optimization
• Deployment on cloud
• Adding more animal classes