Project Overview

Trained a convolutional neural network (CNN) using PyTorch to classify endangered animal species with 82.87% accuracy. Preprocessed a dataset of 6,500 images and implemented data augmentation, dropout, and batch normalization to enhance model generalization and mitigate overfitting. Conducted iterative testing across five model architectures, optimizing hyperparameters and learning rates. Proposed real-world applications for wildlife conservation, including integration with CCTV and social media monitoring for illegal hunting detection.