Quickest product classification according to industry standards (Web and API)

A python framework accelerating ML based discovery in the medical field by encouraging code reuse. Batteries included :)

Deep Learning based Skin Cancer Detection using multiple CNN architectures (VGG, ResNet, DenseNet, EfficientNet, Inception) with image preprocessing using ESRGAN and performance comparison for clinical AI research.

Repository for the paper: "Skin Lesion Classification Using Dermoscopic Images and Clinical Metadata: Insights from Multimodal Models"

ISIC Challenge submission platform.

Instructions for the removal of duplicate image files from within individual ISIC datasets and across all ISIC datasets.

Source code for the paper: "Selective Alignment Transfer for Domain Adaptation in Skin Lesion Analysis".

The official command line tool for interacting with the ISIC Archive.

U-Net & ConvLSTM melanoma segmentation with SLIC superpixels, mask-to-bounding-box ROI cropping, and cross-ISIC (2016–2020) evaluation; ISIC-2017 results: U-Net 86.43% acc / 55.17% IoU, ConvLSTM 85% acc / 48% IoU.

Multimodal melanoma classifier trained on harmonized ISIC 2024 + iToBoS 3D data; metadata-aware pipeline with domain adaptation and cross-dataset evaluation.

Repository for the paper: "Domain Adaptation for Skin Lesion: Evaluating Real-World Generalisation"

Research-driven multi-stage ensemble pipeline for skin cancer detection using deep learning.

This project combines traditional machine learning approaches with advanced deep learning techniques to assist healthcare professionals in early diagnosis and improve patient outcomes.

skin cancer detection model

U-Net-based Models for Skin Lesion Segmentation: More Attention and Augmentation

ISIC Archive API v2 download images by ISIC ID

Source code for the paper: "Dermoscopic Dark Corner Artifacts Removal: Friend or Foe?"

Source code and experiments for the paper: "Dark Corner on Skin Lesion Image Dataset: Does it matter?"

This project aims to predict the presence of skin cancer using a hybrid deep learning model that integrates both tabular data and image data. The dataset used is the ISIC dataset, which contains skin lesion images along with associated metadata. The model achieves an accuracy of 88% on the test set.

A Python SDK Library for working with the International Standard Industrial Classification of All Economic Activities (ISIC), Revision 4.