Deep Learning for Viral Detection: Affordable Camera Technology in Public Health

Viral infections like chickenpox, measles, and monkeypox pose significant global health challenges, affecting millions with varying severity. This study presents a novel deep-learningapproach using widely available low-cost RGB camera technology to accurately identify these infections based on skin manifestations. We aimto enhance diagnostic capabilities and enable timely interventions, thus improving public health outcomes and individual well-being. Using MobileNetV3 for data classification, our model achieved a precision of 95% for positive cases, an overall accuracy of 95.73%, a recall of 88.37%, and an F1-score of 91.56%, indicating balanced performance between precision and recall. Notably, the model demonstrated exceptionally high specificity at 98.34%, effectively identifying negative cases. This deep learning approach holds promise for improving diagnostic accuracy and efficiency, especially in resource-limited settings with limited access to specialized medical expertise. By leveraging low-cost RGB camera technology, our method enables broad deployment, facilitating early detection and treatment of viral infections. We focus onthe potential of deep learning in public health by emphasizing the critical role of early detection and intervention in mitigating the impact of viral infections. Our findings contribute to advancing healthcare technology and lay the groundwork for futureinnovations in disease detection and management.