🦠 Interpretable ViT — Microbial Classifier

Note: This model was trained entirely on procedurally-generated synthetic images (colored shapes with noise standing in for each organism), not real microscopy photographs. It is a demonstration of the interpretable Vision Transformer architecture from our accepted talk at NEXTGEN MICROBIOLOGY II — it will not meaningfully classify real microbial images.

Click to upload an image
(or drag & drop)

Predictions

Upload an image to see predictions here.