This project focuses on building a high-accuracy image classification model to distinguish between cats and dogs using Deep Learning. A pre-trained ResNet50 model with transfer learning is used to achieve excellent performance, reaching 98.6% accuracy.
Kaggle Notebook: Cats vs Dogs Classification on Kaggle
Image classification is a fundamental problem in computer vision. In this project, a Convolutional Neural Network (CNN) based on ResNet50 is fine-tuned to classify images of cats and dogs. The notebook includes data exploration, preprocessing, model training, evaluation, and performance analysis.
Transfer learning allows leveraging knowledge from large-scale datasets (ImageNet) to improve accuracy and reduce training time.
-
Build an accurate image classification model for cats and dogs
-
Apply transfer learning using ResNet50
-
Perform basic exploratory data analysis (EDA)
-
Evaluate model performance using accuracy metrics
-
Achieve high generalization on unseen test data
-
Binary image classification (Cat vs Dog)
-
Transfer learning with ResNet50
-
Image preprocessing and augmentation
-
High accuracy (98.6%)
-
Clean and well-structured training pipeline
-
Model evaluation and performance visualization
- Python
- TensorFlow / Keras
- ResNet50 (Pre-trained on ImageNet)
- NumPy
- Matplotlib
- Seaborn
- Jupyter Notebook
- Dataset Name: Cats vs Dogs Dataset
- Classes:
- Cat
- Dog
- Source: Cats vs Dogs Dataset
-
Model Used: ResNet50 (Transfer Learning)
-
Accuracy Achieved: 98.6%
-
The model demonstrates strong generalization and robust performance on unseen data.
-
EDA:
- Confusion Matrix:
Muqadas Ejaz
BS Computer Science (AI Specialization)
AI/ML Engineer
Data Science & Gen AI Enthusiast
📫 Connect with me on LinkedIn
🌐 GitHub: github.com/muqadasejaz
📬 Kaggle: Kaggle Profile
This project is open-source and available under the MIT License.
⭐ If you find this project useful, don’t forget to star the repository!