Deep Learning in Health Care: Automatic Cervix Image Classification Using Convolutional Neural Network

Cervical malignancy can be viably counteracted if identified in the pre-cancerous stage. In order to appropriately treat cervical cancer, making an accurate determination of a patient’s cervical type is critical. However, doing so can be difficult, even for trained healthcare providers, because of the thin-line difference among the various cervix types. Kaggle and Mobile ODT have distributed a gathering of a few thousand commented on photographs of cervices. In this paper, we utilize profound learning approaches in computer vision, for example, convolutional neural networks and transfer learning. We try different things to experiment the models, for example, batch normalization, image augmentation, and dataset methodologies, for example, cropping the images. The initiations are utilized for the preprepared model Inception v3 which was prepared on the ImageNet dataset of 1.2 million pictures.