Project URL: https://github.com/Shift-Happens/YOLOv3_CatDetector

The YOLOv3_CatDetector project was an ambitious attempt to create a deep learning model capable of accurately detecting cats in images. Using the YOLOv3 (You Only Look Once) object detection model, I trained the system with a dataset of various cat images to help the model recognize feline features in diverse environments. However, things took an unexpected turn during testing. While the model performed well with some images, it soon became clear that YOLOv3_CatDetector had a fuzzy interpretation of "cat." Rather than exclusively detecting cats, the model often flagged any furry object—dogs, pillows, even stuffed animals—as a potential feline. This led to some comical and frustrating misclassifications, revealing that the model's understanding of fur patterns and texture was too broad. The project highlighted the challenges of fine-tuning object detection models to discern subtle differences between visually similar objects, and ultimately, it served as a great learning experience in training data preparation and model refinement.