Probabilistic-MCTrans

Code for Probabilistic Multi-Compound Transformer (Probabilistic MCTrans).

This code are modified and combine code from Probabilistic-Unet-Pytorch and MCTrans. Basically this code add a probabilistic model from Probabilistic U-Net into transformer-based model MCTrans. In previous study, model MCTrans can solved long-range dependencies problem and model Probabilistic U-Net can capture ambiguity in biomedical image. For installation, you can check MCTrans repository.

Probabilistic U-Net

Title: A Probabilistic U-Net for Segmentation of Ambiguous Images.
Authors: Simon A. A. Kohl, Bernardino Romera-Paredes, Clemens Meyer, Jeffrey De Fauw, Joseph R. Ledsam, Klaus H. Maier-Hein, S. M. Ali Eslami, Danilo Jimenez Rezende, Olaf Ronneberger.
[paper]
The architecture of Probabilistic U-Net:
Training process:

Sampling process:

Multi-Compound Transformer (MCTrans)

Title: Multi-Compound Transformer for Accurate Biomedical Image Segmentation
Authors: Yuanfeng Ji, Ruimao Zhang, Huijie Wang, Zhen Li, Lingyun Wu, Shaoting Zhang, Ping Luo.
[paper]
The architecture of MCTranst:
Training process:

Probabilistic Multi-Compound Transformer (Probabilistic MCTrans)

The architecture of Probabilistic MCTrans:
Training process:

Sampling process: