Corroboration of skin diseases: Melanoma, vitiligo vascular tumor using transfer learning

Title

Corroboration of skin diseases: Melanoma, vitiligo vascular tumor using transfer learning

Creator

Agrawal N.; Aurelia S.

Description

The precise identification of skin disease is an exigent process even for more experienced doctors and dermatologists because there is a small variation between surrounding skin and lesions, a visual affinity between different skin diseases. Transfer learning is the approach which stores acquired knowledge while solving one problem and apply that knowledge to similar problems. It is a type of machine learning task where a model proposed for a task can be used again. Transfer learning is used in various areas like image processing and gaming simulation. Image processing is an evolving field in the diagnosis of various kinds of skin diseases. Here transfer learning is used to identify three skin diseases such as melanoma, vitiligo, and vascular tumors. The inception V3 model was used as a base model. Networks were pre-trained and then fine-tuned. Considerable growth of training accuracy and testing accuracy were achieved. 2021 IEEE.

Source

Proceedings of the 7th International Conference on Electrical Energy Systems, ICEES 2021, pp. 590-592.

Date

2021-01-01

Publisher

Institute of Electrical and Electronics Engineers Inc.

Subject

Deep Learning; Melanoma; Skin disease; Transfer Learning; Vascular Tumor; Vitiligo

Coverage

Agrawal N., Christ Deemed to Be University, Department of Computer Science, Bengaluru, India; Aurelia S., Christ Deemed to Be University, Department of Computer Science, Bengaluru, India

Rights

Restricted Access

Relation

ISBN: 978-172817612-3

Format

Online

Language

English

Type

Conference paper

Identifier

https://doi.org/10.1109/ICEES51510.2021.9383682

https://www.scopus.com/inward/record.uri?eid=2-s2.0-85103888624&doi=10.1109%2fICEES51510.2021.9383682&partnerID=40&md5=da610cfe3877253bdd1244f90458f61b