Search Results for author: Stefania Petra

Found 7 papers, 1 papers with code

Self-Certifying Classification by Linearized Deep Assignment

1 code implementation26 Jan 2022 Bastian Boll, Alexander Zeilmann, Stefania Petra, Christoph Schnörr

We propose a novel class of deep stochastic predictors for classifying metric data on graphs within the PAC-Bayes risk certification paradigm.

Classification

Learning Linearized Assignment Flows for Image Labeling

no code implementations2 Aug 2021 Alexander Zeilmann, Stefania Petra, Christoph Schnörr

An exact formula is derived for the parameter gradient of any loss function that is constrained by the linear system of ODEs determining the linearized assignment flow.

Multi-Channel Potts-Based Reconstruction for Multi-Spectral Computed Tomography

no code implementations12 Sep 2020 Lukas Kiefer, Stefania Petra, Martin Storath, Andreas Weinmann

We consider reconstructing multi-channel images from measurements performed by photon-counting and energy-discriminating detectors in the setting of multi-spectral X-ray computed tomography (CT).

Computed Tomography (CT)

Self-Assignment Flows for Unsupervised Data Labeling on Graphs

no code implementations8 Nov 2019 Matthias Zisler, Artjom Zern, Stefania Petra, Christoph Schnörr

This paper extends the recently introduced assignment flow approach for supervised image labeling to unsupervised scenarios where no labels are given.

Combinatorial Optimization

Learning Adaptive Regularization for Image Labeling Using Geometric Assignment

no code implementations22 Oct 2019 Ruben Hühnerbein, Fabrizio Savarino, Stefania Petra, Christoph Schnörr

We study the inverse problem of model parameter learning for pixelwise image labeling, using the linear assignment flow and training data with ground truth.

Numerical Integration

Unsupervised Assignment Flow: Label Learning on Feature Manifolds by Spatially Regularized Geometric Assignment

no code implementations24 Apr 2019 Artjom Zern, Matthias Zisler, Stefania Petra, Christoph Schnörr

Experiments demonstrate a beneficial effect in both directions: adaptivity of labels improves image labeling, and steering label evolution by spatially regularized assignments leads to proper labels, because the assignment flow for supervised labeling is exactly used without any approximation for label learning.

Image Labeling by Assignment

no code implementations16 Mar 2016 Freddie Åström, Stefania Petra, Bernhard Schmitzer, Christoph Schnörr

We introduce a novel geometric approach to the image labeling problem.

Cannot find the paper you are looking for? You can Submit a new open access paper.