A Hierarchical Approach to Skin Lesion Classification

Kulhalli, Rahul; Savadikar, Chinmay; Garware, Bhushan

doi:10.1145/3297001.3297033

Cited by 9 publications

(1 citation statement)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Finally, the (classification) layer also consists of counted stages, and each stage is a layer (flatten, fully connected, softmax). Several researchers have used CNN to classify diseases and especially dermatological diseases by learning to transmit or fine-tune pre-trained models such as Inception v3 [22], [23], [24], Residual neural network [25] Visual Geometry Group architecture, among others. The weights of the original model remain the same, and some optimization classes can be added.…”

Section: -Classification Of Dermatological Diseases Using Deep Learningmentioning

confidence: 99%

A Review Of Skin Cancer Detection

Hussein

Stephan

2021

J. Al-Qadisiyah Comp. Sci. Math.

View full text Add to dashboard Cite

It can be said that the most common or common health disease is a skin disease. skin diseases are often determined by doctors' experience and sample results (skin biopsy), and it is certainly a time-consuming process. Therefore, there has become an urgent need for an automated system (computer) to identify and discover skin diseases through images with very high accuracy, with fewer doctors or experts in this purview. Identification and classification of the skin disease through the feature/s and characteristics that were taken from these images. Since skin diseases have very similar optical properties And therefore add a lot more challenges to choosing the useful feature/s of the image. This means that the accurate analysis of these skin diseases through images will have a good prognosis, short diagnostic time, and speed in diagnosis, and thus it will facilitate and cost-effective treatment. This paper provides an overview or study on the different methods and techniques for identifying and classifying skin diseases, which are traditional technology and technology based on deep learning.

show abstract