One of the most prevalent dental diseases globally is dental caries. It is the medical name for the widespread condition known as tooth decay or cavities. People of all ages are susceptible to dental decay these days. Immediate diagnosis is crucial for effective treatment to prevent the patient from suffering from potentially life-threatening consequences. Dental caries, also referred to as tooth decay, is a widespread chronic condition that occurs when tooth enamel breaks down. The acid produced by the bacteria in the mouth causes tooth decay by eroding the enamel and eventually leaving a tiny hole in the tooth. If left untreated, oral illnesses may cause a cascade of complications, including receding gums, cavities, tooth loss, and even bone loss. The dentist greatly benefits from X-ray scans. To help with patient diagnosis and treatment, they use these images. On the other hand, not even a trained eye can easily decipher X-ray images. Many in the medical industry are interested in tooth recognition and teeth separation from X-ray images due to its practical applications. The objective here is to categorize illness rather than the progression through its phases, but there may be many stages of dental caries.In this study, we classified dental images to identify caries using Nesterov Accelerated Gradient (NAG) optimization using EfficientNet-B0. For the purpose of dental carrier segmentation, we used an improved Unet with ANN. When we compared our approach to others, the findings showed that it was more accurate and performed better.