ArXiv

Authors Archit A, Freckmann L, Pape C Journal Arxiv Citation arXiv:2501.11734. Abstract Medical image segmentation is an important analysis task in clinical practice and research. Deep learning has massively advanced the field, but current approaches are mostly based on models trained for a specific task. Training such models or adapting

Authors Engelken R, Monteforte M, Wolf F   Journal Arxiv   Citation arXiv:2412.21188.   Abstract Nerve impulses, the currency of information flow in the brain, are generated by an instability of the neuronal membrane potential dynamics. Neuronal circuits exhibit collective chaos that appears essential for learning, memory, sensory processing, and

Authors Schmidt A, Hiemeyer P, Wolf F Journal Arxiv Citation arXiv:2305.13420. Abstract Distributions of neuronal activity within cortical circuits are often found to display highly skewed shapes with many neurons emitting action potentials at low or vanishing rates, while some are active at high rates. Theoretical studies were able to

Authors Schneider AC, Neuhaus V, Ehrlich DA, Makkeh A, Ecker AS, Priesemann V, Wibral M Journal Arxiv Citation arXiv:2412.02482. Abstract In modern deep neural networks, the learning dynamics of the individual neurons is often obscure, as the networks are trained via global optimization. Conversely, biological systems build on self-organized, local

Authors Gonualez-Sanz A, Groppe M, Munk A Journal Arxiv Citation arXiv:2410.23146. Abstract The goal of optimal transport (OT) is to find optimal assignments or matchings between data sets which minimize the total cost for a given cost function. However, sometimes the cost function is unknown but we have access to

Authors Suchan L, Li H, Munk A Journal Arxiv Citation arXiv:2407.15297. Abstract Graph cuts are among the most prominent tools for clustering and classification analysis. While intensively studied from geometric and algorithmic perspectives, graph cut-based statistical inference still remains elusive to a certain extent. Distributional limits are fundamental in understanding

Authors Milosevic N, Müller G, Huisken J, Scherf N Journal Arxiv Citation arXiv:2406.03845. Abstract In this work, we introduce the concept of Active Representation Learning, a novel class of problems that intertwines exploration and representation learning within partially observable environments. We extend ideas from Active Simultaneous Localization and Mapping (active

Authors Liu Z, Li H   Journal Arxiv   Citation arXiv:2403.11356.   Abstract For robust and efficient detection of change points, we introduce a novel methodology MUSCLE (multiscale quantile segmentation controlling local error) that partitions serial data into multiple segments, each sharing a common quantile. It leverages multiple tests for

Authors Wang X, Scholand N, Tan Z, Mackner D, Telezki V, Blumenthal M, Schaten P, Uecker M Journal Arxiv Citation arXiv:2402.05366. Abstract Purpose: To develop a model-based nonlinear reconstruction for simultaneous water-specific T1, R∗2, B0 field and/or fat fraction (FF) mapping using single-shot inversion-recovery (IR) multi-echo radial FLASH. Methods: The

Authors Engelken R Journal Arxiv Citation arXiv:2312.17216. Abstract Spiking Neural Networks (SNNs) are biologically-inspired models that are capable of processing information in streams of action potentials. However, simulating and training SNNs is computationally expensive due to the need to solve large systems of coupled differential equations. In this paper, we