The first polyp segmentation dataset with self-generated annotations

Large size

Synth-Colon is made of 20.000 synthetic images with 500x500 resolution.

Realisitc

The dataset includes realistic images generated using CycleGAN and the Kvasir dataset.

3D mesh

3D mesh of the colon and polyps in OBJ format.

Depth map

Detailed depth information in OpenEXR format.


    We release CUT-seg, a new model that is computationally less expensive, and requires less real images than CycleGAN. It is composed by a segmentation model and a generative model that are jointly trained to produce realistic images while learning to segment polyps. We take advantage of recent one-sided translation models because they use significantly less memory, allowing us to add a segmentation model in the training loop.

  • Source code