- Developed a CNN-based deep learning model to categorise dog breeds, utilising the Stanford Dogs Dataset for preprocessing, training, and evaluation
- Implemented an optimised CNN architecture, leveraging Xception with 71 layers, registering 83% accuracy rate