Implementasi YOLOv3 Menggunakan Fitur Ekstraktor ResNeXt Untuk Deteksi Filariasis

Filariasis is a serious infectious disease prevalent in many tropical and subtropical countries, including Indonesia, caused by infection from filarial worms transmitted through mosquito bites. The disease can lead to severe limb swelling (elephantiasis) and permanent disability if not treated promptly. Early detection of Filariasis is crucial in preventing the progression of serious illness, reducing transmission, and lowering long-term treatment costs. Vector control and mass treatment with antiparasitic drugs are the main prevention strategies, while recent research explores AI-based detection methods, such as Convolutional Neural Networks (CNN), Faster R-CNN, K-Nearest Neighbors (KNN), SSD, and YOLO, to improve diagnostic efficiency for Filariasis. This study aims to integrate YOLOv3 and ResNeXt algorithms into a Filariasis detection model, with the hope of producing a fast, accurate, and efficient approach to identifying this disease. As a result, AI technology has the potential to support efforts in eradicating Filariasis through more effective and accurate early detection. The implementation of the YOLOv3 algorithm with ResNeXt as a feature extractor in object detection processes demonstrated excellent performance, with an average accuracy of 96.77%. This indicates that the object detection model is reliable and suitable for use in systems requiring accurate object detection, particularly in various types of anomalies and medical images.