SkinSage - Lesion Diagnosis using Deep Learning Techniques

Authors

Nimish
Dept. of CSE, Amity University, Uttar Pradesh, Noida Campus
Paarth Bharwaj
Dept. of CSE, Amity University, Uttar Pradesh, Noida Campus
Vaibhav Bhargava
Dept. of CSE, Amity University, Uttar Pradesh, Noida Campus
Shivendra Singh
Dept. of CSE, Amity University, Uttar Pradesh, Noida Campus

Synopsis

The alarming rise in instances of skin cancer in recent years, one of the most prevalent malignancies worldwide, emphasises how important early and accurate identification is. The SkinSage project, which combines the precision of deep learning. In the past, dermatologists relied on eye exams to spot skin lesions, with the use of tools like dermoscopy on occasion. But with datasets like HAM10000, digital photography has been able to record a broad variety of skin conditions, creating a rich tapestry of data that is available for computer analysis. SkinSage uses cutting-edge neural architectures like ResNet, DenseNet, and InceptionV3 to recognise patterns and nuances that are much above the capabilities of the human eye. In addition to being a computational marvel, the SkinSage project also exemplifies accessibility by making its analytical capabilities available via a smartphone application. High-quality skin lesion diagnostics are now accessible to everyone thanks to this platform, which was developed with an unwavering emphasis on the user experience, regardless of location or degree of training. With the confluence of these developments, SkinSage becomes more than just a tool for diagnosing skin cancer; it signals a paradigm shift that has the potential to democratise early detection and save countless lives in the future.

RAMSA 2024
Published
February 29, 2024