This project identifies fruits from a live camera feed, marking each fruit with its name and detection score by drawing bounding boxes around them.

At present, the dataset includes 1745 images spanning 4 classes, resulting in approximately 5000 instances. Training was conducted using YOLOv9, and the model is available as the "best.pt" file.

The video showcasing the project can be accessed through the link provided by clicking on the image below.