Skip to content

Final project for "Bayesian machine learning" course @ Skoltech

Notifications You must be signed in to change notification settings

RodinIvan/Probabilistic_UNet_BML

Repository files navigation

Probabilistic_UNet_BML

Final project for "Bayesian machine learning" course @ Skoltech

  1. A Probabilistic U-Net for Segmentation of Ambiguous Images: https://arxiv.org/pdf/1806.05034.pdf

  2. Many real-world vision problems suffer from inherent ambiguities. In clinical applications for example, it might not be clear from a CT scan alone which particular region is cancer tissue. Therefore a group of graders typically produces a set of diverse but plausible segmentations. We consider the task of learning a distribution over segmentations given an input. To this end we propose a generative segmentation model based on a combination of a U-Net with a conditional variational autoencoder that is capable of efficiently producing an unlimited number of plausible hypotheses. We show on a lung abnormalities segmentation task and on a Cityscapes segmentation task that our model reproduces the possible segmentation variants as well as the frequencies with which they occur, doing so significantly better than published approaches. These models could have a high impact in real-world applications, such as being used as clinical decision-making algorithms accounting for multiple plausible semantic segmentation hypotheses to provide possible diagnoses and recommend further actions to resolve the present ambiguities.

  3. Contributors: Ivan Rodin, Vladislav Shlyonskiy

  4. Dataset used at project: https://competitions.codalab.org/competitions/17094

About

Final project for "Bayesian machine learning" course @ Skoltech

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages