A Deep Learning Approach – Monkey Detection using YOLOv7
M Vamshi, Navateja Marupaka, Sri Charan Nallamothu, Ayesha Naureen · 2023
This study develops a cutting-edge computer vision system for monkey recognition using YOLOv7. Our unique dataset consists of monkey images from natural habitats and personal collections, divided into training and testing phases. Manual annotation was done with bounding boxes using tools like labelImg. The YOLOv7 model was trained with a fixed schedule and hyperparameters optimized for monkey detection. The model went through specific epoch training, and we selected the checkpoint with the best performance for analysis. Performance metrics included accuracy, recall, mAP, [email protected] and F1-score, all of which exhibited distinct values of 0.995, 0.995, 0.993, 0.996, and 0.99, respectively. These results show that remarkable adaptation of the model has been demonstrated to a variety of animal recognition tasks, including YOLOv7 transformation and conventional datasets, indicating widespread use. In summary, this study successfully uses YOLOv7 for monkey recognition and demonstrates its potential in a variety of object detection applications.