cat-dog-classifier

🐶🐱 Cat vs Dog Classifier (FastAI + ResNet34)

This project uses FastAI and a pretrained ResNet34 model to classify cats vs dogs.
Dataset: Oxford-IIIT Pets (subset used: cats & dogs).

---

🚀 Project Overview

• Built an image classifier with transfer learning (ResNet34 backbone).
• Achieved ~90%+ validation accuracy.
• Performed error analysis with confusion matrix and top losses visualization.
• Packaged into a reproducible Jupyter notebook.

---

📊 Results

Validation Accuracy: ~90%+

Confusion Matrix

Top Losses (Misclassified Examples)

Sample Prediction

---

📂 Files

• cat_dog_classifier.ipynb → Training & evaluation notebook
• results/ → Visualizations (confusion matrix, top losses, sample prediction)

---

⚙️ How to Run

pip install fastai torch torchvision matplotlib pillow