Efficient Real-time Breed Classification using YOLOv7 Object Detection Algorithm

This work aims to develop an accurate and reliable solution for real-time breed recognition and categorisation of dogs and cats using the YOLO-v7 object detection algorithm. The Oxford IIIT pet dataset, consisting of a large number of images of different dog and cat breeds with various poses and backgrounds, is utilized for training and validation of the model. This model achieved a high F1 score of 85.6% and mAP of 82.57%, superior than other latest models. The findings illustrate the potential of using the YOLO-v7 algorithm for real-time breed detection and classification of dogs and cats, which can be used in a wide range of fields, including veterinary clinics, animal shelters, and pet stores.