Determining the Efficacy of SENet Integrated YOLO Models For Animal Detection

Monitoring wildlife is essential for conservation efforts, but it can be difficult to find and identify species in a variety of settings in real time. Many existing models struggle with low light, occluded and far away images. In this study, we provide a multiclass animal detection system that is integrated with the state-of-the-art artificial intelligence framework. Using information from our customized dataset, the system uses sophisticated detection models to identify and categorize a variety of animal species in diverse settings. The system provides precise animal detection and improved accuracy of 90% on detecting small animals by integrating SENet in YOLOv7 and YOLOv8. The data processing capabilities in YOLOv8 is higher compared to YOLOv7. Additionally, the system provides useful information for tracking wildlife, supporting conservation initiatives and lowering confrontations between people and animals.