Skin Cancer ViT Predictor
This application demonstrates the SkinCancerViT multimodal model for skin cancer diagnosis.
It can take an uploaded image with patient metadata or predict on a random sample from the dataset.
It also visualizes the saliency map generated by the Vision Transformer component.
Disclaimer: This tool is for demonstration and research purposes only and should not be used for medical diagnosis.
Get a Prediction from a Random Sample in the Test Set
Random sample details and prediction will appear here.