Neural Network Based Classification of Dental Diseases Using Augmentation Techniques

Abstract

Due to the advancement in deep learning models, they have established a strong foothold in the field of medicine. Dentists are relying increasingly on classification models that employ deep learning. Deep neural networks (DNNs) are the most effective and prevalent machine learning models used for a variety of image analysis tasks including image retrieval, image classi fication, object detection, and 3D analysis. They have reached performance levels comparable to those of humans. DNNs are becoming increasingly com mon for roles like medical image processing, landmark/organ, localization, cancer diagnosis, diabetic retinopathy detection, Covid-19 identification, and teeth classification due to their success with natural images (such as images captured from natural scenes, such as imagenet and CIFAR-10). Despite the increase in the development of wise intelligent systems through the incorpo ration of deep structured networks, the vulnerability of such systems may be compromised by the mere presence of an adversarial perturbation. In this research, a new approach is proposed that will classify different teeth diseases and also to improve accuracy the GANs has been implemented.