Source author record

Vedrana Andersen Dahl

Vedrana Andersen Dahl appears in the imported research catalog. Authorship, coauthor and topic links are available while profile ownership is still unclaimed.

ResearcherUnclaimed source record

Computer Vision cond-mat.mtrl-sci Data Structures and Algorithms eess.IV

Catalog footprint

What is connected

3works

4topics

4close collaborators

Actions

Connect this record

Open graph Browse works

Inspect adjacent papers, topics, institutions and collaborators without losing the researcher page.

Building this map preview

BZPEER is loading the nearby papers, people, topics and institutions for this page.

preprint2026arXiv

Fast and Compact Graph Cuts for the Boykov-Kolmogorov Algorithm

Computing a minimum $s$-$t$ cut in a graph is a solution to a wide range of computer vision problems, and is often done using the Boykov-Kolmogorov (BK) algorithm. In this paper, we revisit the BK algorithm from both a theoretical and practical point of view. We improve the analysis of the time complexity of the BK algorithm to $O(mn|C|)$ and propose a new algorithm, the fast and compact BK (fcBK) algorithm, with a time complexity of $O(m|C|)$, where $m$, $n$, and $|C|$ are the number of edges, number of vertices, and the capacity of the cut, respectively. We additionally propose a compact graph representation that allows our implementation to find a minimum $s$-$t$ cut in a graph with upwards of $10^9$ vertices and $10^{10}$ edges on a machine with 128 GB of memory. We find our implementation of the BK algorithm to be the fastest available implementation of the BK algorithm when evaluating on a comprehensive set of benchmark datasets, highlighting the importance of memory-efficient implementations. We make our implementations publicly available for further research and implementation development within minimum $s$-$t$ cut algorithms.

preprint2026arXiv

Towards Agnostic and Holistic Universal Image Segmentation with Bit Diffusion

This paper introduces a diffusion-based framework for universal image segmentation, making agnostic segmentation possible without depending on mask-based frameworks and instead predicting the full segmentation in a holistic manner. We present several key adaptations to diffusion models, which are important in this discrete setting. Notably, we show that a location-aware palette with our 2D gray code ordering improves performance. Adding a final tanh activation function is crucial for discrete data. On optimizing diffusion parameters, the sigmoid loss weighting consistently outperforms alternatives, regardless of the prediction type used, and we settle on x-prediction. While our current model does not yet surpass leading mask-based architectures, it narrows the performance gap and introduces unique capabilities, such as principled ambiguity modeling, that these models lack. All models were trained from scratch, and we believe that combining our proposed improvements with large-scale pretraining or promptable conditioning could lead to competitive models.

preprint2021arXiv

A Physical Model for Microstructural Characterization and Segmentation of 3D Tomography Data

We present a novel method for characterizing the microstructure of a material from volumetric datasets such as 3D image data from computed tomography (CT). The method is based on a new statistical model for the distribution of voxel intensities and gradient magnitudes, incorporating prior knowledge about the physical nature of the imaging process. It allows for direct quantification of parameters of the imaged sample like volume fractions, interface areas and material density, and parameters related to the imaging process like image resolution and noise levels. Existing methods for characterization from 3D images often require segmentation of the data, a procedure where each voxel is labeled according to the best guess of which material it represents. Through our approach, the segmentation step is circumvented so that errors and computational costs related to this part of the image processing pipeline are avoided. Instead, the material parameters are quantified through their known relation to parameters of our model which is fitted directly to the raw, unsegmented data. We present an automated model fitting procedure that gives reproducible results without human bias and enables automatic analysis of large sets of tomograms. For more complex structure analysis questions, a segmentation is still beneficial. We show that our model can be used as input to existing probabilistic methods, providing a segmentation that is based on the physics of the imaged sample. Because our model accounts for mixed-material voxels stemming from blurring inherent to the imaging technique, we reduce the errors that other methods can create at interfaces between materials.