Skin lesion classification: Datasets & Models
A computer vision workflow that classifies dermatoscopic and clinical skin images into lesion categories such as melanoma, nevus, and benign keratosis. Built to support skin lesion detection and triage in healthcare imaging pipelines.
Skin lesion classification
This workflow performs skin lesion detection by classifying dermatoscopic and clinical skin photographs into categories such as melanoma, basal cell carcinoma, benign nevus, and seborrheic keratosis. It takes single skin images as input and outputs a predicted lesion class based on visual features like pigmentation, border shape, texture, and color variation.
Try it out
Powered by general detection model
Run on custom image
Drop an image here or click to upload
Why automate skin lesion classification
Manual skin lesion review is time-intensive and depends on dermatologist availability, which creates backlogs and inconsistent assessments across clinics and reviewers. Automating classification with computer vision gives builders a repeatable, high-throughput first pass that flags suspicious lesions for expedited human review and reduces the volume of clearly benign cases that consume specialist time. In a healthcare imaging setting, this means faster triage, more consistent labeling of lesion types, and a scalable way to pre-screen large image archives. A ViT-based classifier is well suited here because it captures both fine local detail and global lesion structure that traditional methods often miss.
Key benefits
Teams get consistent, per-image lesion classifications that integrate with existing dermatoscopy hardware, PACS archives, and telemedicine intake workflows without new capture infrastructure. The Vision Transformer architecture scales from single-clinic deployments to large research datasets while maintaining reproducible outputs. Because it runs on standard image inputs, the workflow slots into both real-time triage tools and batch processing of historical imagery.
How to use it
- Upload an image: Drop any photo into the widget above to try skin lesion classification for free, right on this page.
- Preview instant results: The model detects melanoma, benign nevus, basal cell carcinoma, seborrheic keratosis, actinic keratosis within seconds and returns an annotated image with structured predictions, so you can immediately see how well it works. Adjust the detection classes and re-run until the results fit your scenario.
- Like what you see?: With a free Roboflow account, you can deploy this workflow with the Serverless Cloud API, customize it in your workspace, or train Classification (ViT) on your own images for production accuracy.
- Get started for free: No code or training data required to begin: upload, review the results, and make it part of your stack in a single click.
Recommended model for skin lesion classification
When you're ready to build a production model for skin lesion classification, we recommend training Classification (ViT). Unlike the ready-to-use model in the demo above, you fine-tune it on your own labeled images for higher accuracy, then deploy to a Serverless Cloud API endpoint.
Not sure where to start?
Describe your Skin Lesion Classification project to Roboflow Agent.