PERBANDINGAN MODEL TRANSFER LEARNING CNN PADA KLASIFIKASI CITRA PENYAKIT GIGI DAN LIDAH

The mouth is one of the parameters that determine the health of the body. In the mouth there are teeth and tongue which are very vulnerable to various diseases. Previously, many studies have done classification for teeth and tongue using the CNN algorithm. However, there is no research that classifies by combining dental and tongue disease datasets. Therefore, this research focuses on the classification of dental and tongue diseases with transfer learning techniques using CNN architecture models VGG16, VGG19, ResNet50. The main purpose of this research is to compare the three models to find the most optimal model performance results in handling related cases. Based on the results of the research that has been done, the best accuracy is obtained in scenarios with data that has been augmented and models trained using 75 epochs. Where the VGG16 model achieved 94% accuracy, VGG19 with 93% accuracy, and ResNet50 accuracy reached 94%.