CNN Wildfire Detection

This WIldfire project analyzes the performance tradeoffs between accuracy and speed through CNN-based experiments.

The YOLOv7 and Tiny YOLOv7 models were implemented for highly nuanced object detection.

Model training, inference evaluations (CPU vs. GPU), and parameter tuning (image sizes, epoch counts) were showcased to efficiently detect fire and smoke in images.

I utilized the PyTorch Python library in Google Colab for this project.