Institution profile

Zurich University of Applied Sciences

Academic institutioneurope · ch
Official website
Research library135linked papers
Opportunities0open roles
Selected work

Representative Papers

Benchmarking the CoW with the TopCoW Challenge: Topology-Aware Anatomical Segmentation of the Circle of Willis for CTA and MRA

Dec 29, 2023arXiv.org

The Circle of Willis (CoW) suffers from a scarcity of high-quality voxel-level annotations in CTA/MRA imaging, reliance on labor-intensive expert manual segmentation, and poor guarantee of topological consistency. Method: We introduce the first publicly available voxel-level multi-class CoW dataset—comprising 13 vascular structures with paired MRA/CTA volumes—and propose a topology-aware segmentation framework: (i) a novel VR-assisted annotation paradigm ensuring anatomical plausibility; (ii) a multimodal registration and topology-constrained segmentation network; and (iii) topology-sensitive metrics including branch F1 and topo-Dice. Contribution/Results: This benchmark has attracted >140 teams across four continents. State-of-the-art models achieve ≈90% Dice on most arterial branches, while exposing persistent topological matching bottlenecks—particularly for communicating arteries and anatomical variants.

24 citations2 influentialRead paper

ROSBag MCP Server: Analyzing Robot Data with LLMs for Agentic Embodied AI Applications

Nov 05, 2025

Current embodied AI research lacks systematic tooling for synergistic analysis of multimodal robotic data—such as trajectories, LiDAR scans, and time-series sensor streams—with large language models (LLMs) or vision-language models (VLMs). Method: We introduce the first Model Context Protocol (MCP)-based ROS/ROS 2 robotics data analytics server, tightly integrating domain-specific knowledge to build a dedicated toolchain. It enables natural-language-driven querying, visualization, and processing—including `ros2 bag` management, topic filtering, and temporal cropping—alongside a lightweight UI for cross-model LLM/VLM benchmarking. Contribution/Results: We evaluate eight state-of-the-art models and find Kimi K2 and Claude Sonnet 4 achieve the highest tool-calling success rates. Our analysis identifies tool description quality, parameter scale, and toolset complexity as key determinants of performance. This work establishes a scalable, interpretable, and model-agnostic analytics paradigm for agentic embodied AI.

1 citationsRead paper

Critical Dynamics of Random Surfaces and Multifractal Scaling

May 29, 2025

This study investigates the temporal evolution of order parameters in conformal field theories (CFTs) at criticality on random surfaces, moving beyond conventional fixed-topology constraints (e.g., fixed area or genus) to explore intrinsic stochastic geometric effects. Method: We develop, for the first time, a multifractal analytical framework coupling random geometry with CFT, integrating critical phenomena theory and multifractal scaling laws to rigorously derive the Hurst exponent spectrum. Contribution/Results: We establish universal multifractal structure across the Ising model, three-state Potts model, and general minimal models. Crucially, higher-order temporal variation scaling laws quantitatively reproduce the multifractal signatures observed in real financial time series. This work constructs a computationally tractable bridge between statistical physics critical systems and complex financial dynamics, offering a novel theoretical paradigm and empirical toolkit for cross-scale complex systems.

1 citationsRead paper

Trends and Reversion in Financial Markets on Time Scales from Minutes to Decades

Jan 28, 2025

This study systematically investigates trend and mean-reversion behaviors in financial markets across time scales ranging from minutes to centuries. Prior research lacks a unified empirical characterization of how market dynamics shift between trending and reverting regimes as a function of observation horizon. Method: Leveraging multi-asset, multi-frequency data (tick-level to annual) spanning equities, interest rates, FX, and commodities, we apply statistical significance testing, duration modeling, and analogies to critical phenomena. Contribution/Results: We provide the first empirical validation that trending and mean-reverting regimes alternate with scale: strong trends dominate at hourly to multi-year horizons, whereas mean reversion prevails at sub-minute and decadal-plus scales; a critical transition zone emerges at 1-hour–several-days, exhibiting memory persistence up to several years. We further propose and validate a novel lattice-gas market model grounded in social network theory, demonstrating that weak trends persist under trending regimes while strong trends trigger reversal, and vice versa under mean-reverting regimes.

1 citationsRead paper
Recent publications

Latest Papers