arXiv subjects

Michael Barz

Publications and source records attributed to Michael Barz.

14 recordsLinked to original sources

arXiv2026

Non-Abelian p-Curvature and a Non-Abelian Katz's Formula

Let $k$ be a field of characteristic $p,$ and $f : X \to S$ a smooth proper morphism of smooth $k$-schemes. Katz's formula gives a relationship between the Kodaira--Spencer map of $f,$ and an invariant called the $p$-curvature of the Gauss--Manin connection associated to $f.$ Recently, Lam--Litt proved a variant of Katz's formula in non-abelian Hodge theory, and suggested that it should be possible to give a more conceptual proof of their formula using the stacky approach to $p$-adic Hodge theory. In this article, we realize their suggestion, explaining how the rather concrete phenomena observed by Katz and Lam--Litt can be explained in a conceptual way using sheared de Rham stacks, as developed by Bhatt--Kanaev--Vologodsky--Zhang and Drinfeld (though we prove a slightly different statement than Lam--Litt do). We do not assume the reader has any background in the theory of de Rham stacks.

math.AG↗

arXiv2026

StructSAM: Structure- and Spectrum-Preserving Token Merging for Segment Anything Models

Recent token merging techniques for Vision Transformers (ViTs) provide substantial speedups by reducing the number of tokens processed by self-attention, often without retraining. However, their direct application to the Segment Anything Model (SAM) family is nontrivial: SAM's image encoder mixes windowed and global attention, and its mask decoder relies on dense, prompt-conditioned features for precise boundary prediction. We systematically evaluate representative token-merging methods on SAM and Medical SAM in a strict off-the-shelf setting, and find that existing destination-selection heuristics can erode boundaries and leak prompt information as merge rates increase. We propose \textbf{StructSAM}, a resolution-preserving merge-unmerge framework tailored to SAM. StructSAM computes a lightweight token-energy score from first-order feature gradients, uses grid-based flatness screening to protect boundary and prompt regions, and merges tokens within flat areas toward low-energy destinations with explicit token recovery. We further provide a spectral graph coarsening view showing that score-guided merging yields bounded Laplacian spectral distortion compared to random or window-restricted baselines. Across eight natural and medical benchmarks, StructSAM reduces encoder FLOPs by 25-30\% (up to 40\%+ with prompt-aware merging) with minor drops in mIoU/Dice, consistently outperforming ToMe, PiToMe, ToMeSD, VidToMe, and ALGM at the same compute.

cs.CV↗

arXiv2026

Heads Up!: Towards In Situ Photogrammetry Annotations and Augmented Reality Visualizations for Guided Backcountry Skiing

Backcountry skiing is an activity where a group of skiers navigate challenging environmental conditions to ski outside of managed areas. This activity requires careful monitoring and effective communication around the current weather and terrain conditions to ensure skier safety. We aim to support and facilitate this communication by providing backcountry guides with a set of in situ spatial annotation tools to communicate hazards and appropriate speeds to the ski recreationalists. A guide can use a tablet application to annotate a photogrammetry-based map of a mountainside, for example, one collected using a commercial camera drone, with hazard points, slow-down zones, and safe zones. These annotations are communicated to the skiers via visual overlays in augmented reality heads-up displays. We present a prototype consisting of a web application and a virtual reality display that mirror the guide's and skier's perspectives, enabling participatory interaction design studies in a safe environment.

cs.HC↗

arXiv2025

Logarithmic de Rham Stacks and Non-Abelian Hodge Theory

In this article, we introduce the logarithmic de Rham stack of a pair (X, D), for a smooth variety X over a field k of positive characteristic p, and D a strict normal crossings divisor on X. Using this stack, we prove a new version of logarithmic Cartier descent, and a new logarithmic non-abelian Hodge theorem for curves, both stated using a certain logarithmic Frobenius twist. Our logarithmic non-abelian Hodge theorem implies an earlier logarithmic non-abelian Hodge theorem of de Cataldo-Zhang.

math.AG↗

arXiv2025

How Many Tokens Do 3D Point Cloud Transformer Architectures Really Need?

Recent advances in 3D point cloud transformers have led to state-of-the-art results in tasks such as semantic segmentation and reconstruction. However, these models typically rely on dense token representations, incurring high computational and memory costs during training and inference. In this work, we present the finding that tokens are remarkably redundant, leading to substantial inefficiency. We introduce gitmerge3D, a globally informed graph token merging method that can reduce the token count by up to 90-95% while maintaining competitive performance. This finding challenges the prevailing assumption that more tokens inherently yield better performance and highlights that many current models are over-tokenized and under-optimized for scalability. We validate our method across multiple 3D vision tasks and show consistent improvements in computational efficiency. This work is the first to assess redundancy in large-scale 3D transformer models, providing insights into the development of more efficient 3D foundation architectures. Our code and checkpoints are publicly available at https://gitmerge3d.github.io

cs.CV↗

arXiv2025

Your Interface, Your Control: Adapting Takeover Requests for Seamless Handover in Semi-Autonomous Vehicles

With the automotive industry transitioning towards conditionally automated driving, takeover warning systems are crucial for ensuring safe collaborative driving between users and semi-automated vehicles. However, previous work has focused on static warning systems that do not accommodate different driver states. Therefore, we propose an adaptive takeover warning system that is personalised to drivers, enhancing their experience and safety. We conducted two user studies investigating semi-autonomous driving scenarios in rural and urban environments while participants performed non-driving-related tasks such as text entry and visual search. We investigated the effects of varying time budgets and head-up versus head-down displays for takeover requests on drivers' situational awareness and mental state. Through our statistical and clustering analyses, we propose strategies for designing adaptable takeover systems, e.g., using longer time budgets and head-up displays for non-hazardous takeover events in high-complexity environments while using shorter time budgets and head-down displays for hazardous events in low-complexity environments.

cs.HC↗

arXiv2025

InFL-UX: A Toolkit for Web-Based Interactive Federated Learning

This paper presents InFL-UX, an interactive, proof-of-concept browser-based Federated Learning (FL) toolkit designed to integrate user contributions seamlessly into the machine learning (ML) workflow. InFL-UX enables users across multiple devices to upload datasets, define classes, and collaboratively train classification models directly in the browser using modern web technologies. Unlike traditional FL toolkits, which often focus on backend simulations, InFL-UX provides a simple user interface for researchers to explore how users interact with and contribute to FL systems in real-world, interactive settings. By prioritising usability and decentralised model training, InFL-UX bridges the gap between FL and Interactive Machine Learning (IML), empowering non-technical users to actively participate in ML classification tasks.

cs.LG↗

arXiv2024

A look under the hood of the Interactive Deep Learning Enterprise (No-IDLE)

This DFKI technical report presents the anatomy of the No-IDLE prototype system (funded by the German Federal Ministry of Education and Research) that provides not only basic and fundamental research in interactive machine learning, but also reveals deeper insights into users' behaviours, needs, and goals. Machine learning and deep learning should become accessible to millions of end users. No-IDLE's goals and scienfific challenges centre around the desire to increase the reach of interactive deep learning solutions for non-experts in machine learning. One of the key innovations described in this technical report is a methodology for interactive machine learning combined with multimodal interaction which will become central when we start interacting with semi-intelligent machines in the upcoming area of neural networks and large language models.

cs.LG↗

arXiv2024

I-MPN: Inductive Message Passing Network for Efficient Human-in-the-Loop Annotation of Mobile Eye Tracking Data

Comprehending how humans process visual information in dynamic settings is crucial for psychology and designing user-centered interactions. While mobile eye-tracking systems combining egocentric video and gaze signals can offer valuable insights, manual analysis of these recordings is time-intensive. In this work, we present a novel human-centered learning algorithm designed for automated object recognition within mobile eye-tracking settings. Our approach seamlessly integrates an object detector with a spatial relation-aware inductive message-passing network (I-MPN), harnessing node profile information and capturing object correlations. Such mechanisms enable us to learn embedding functions capable of generalizing to new object angle views, facilitating rapid adaptation and efficient reasoning in dynamic contexts as users navigate their environment. Through experiments conducted on three distinct video sequences, our interactive-based method showcases significant performance improvements over fixed training/testing algorithms, even when trained on considerably smaller annotated samples collected through user feedback. Furthermore, we demonstrate exceptional efficiency in data annotation processes and surpass prior interactive methods that use complete object detectors, combine detectors with convolutional networks, or employ interactive video segmentation.

cs.CV↗

arXiv2022

DRG-Net: Interactive Joint Learning of Multi-lesion Segmentation and Classification for Diabetic Retinopathy Grading

Diabetic Retinopathy (DR) is a leading cause of vision loss in the world, and early DR detection is necessary to prevent vision loss and support an appropriate treatment. In this work, we leverage interactive machine learning and introduce a joint learning framework, termed DRG-Net, to effectively learn both disease grading and multi-lesion segmentation. Our DRG-Net consists of two modules: (i) DRG-AI-System to classify DR Grading, localize lesion areas, and provide visual explanations; (ii) DRG-Expert-Interaction to receive feedback from user-expert and improve the DRG-AI-System. To deal with sparse data, we utilize transfer learning mechanisms to extract invariant feature representations by using Wasserstein distance and adversarial learning-based entropy minimization. Besides, we propose a novel attention strategy at both low- and high-level features to automatically select the most significant lesion information and provide explainable properties. In terms of human interaction, we further develop DRG-Net as a tool that enables expert users to correct the system's predictions, which may then be used to update the system as a whole. Moreover, thanks to the attention mechanism and loss functions constraint between lesion features and classification features, our approach can be robust given a certain level of noise in the feedback of users. We have benchmarked DRG-Net on the two largest DR datasets, i.e., IDRID and FGADR, and compared it to various state-of-the-art deep learning networks. In addition to outperforming other SOTA approaches, DRG-Net is effectively updated using user feedback, even in a weakly-supervised manner.

cs.CV↗

arXiv2019

Incremental Improvement of a Question Answering System by Re-ranking Answer Candidates using Machine Learning

We implement a method for re-ranking top-10 results of a state-of-the-art question answering (QA) system. The goal of our re-ranking approach is to improve the answer selection given the user question and the top-10 candidates. We focus on improving deployed QA systems that do not allow re-training or re-training comes at a high cost. Our re-ranking approach learns a similarity function using n-gram based features using the query, the answer and the initial system confidence as input. Our contributions are: (1) we generate a QA training corpus starting from 877 answers from the customer care domain of T-Mobile Austria, (2) we implement a state-of-the-art QA pipeline using neural sentence embeddings that encode queries in the same space than the answer index, and (3) we evaluate the QA pipeline and our re-ranking approach using a separately provided test set. The test set can be considered to be available after deployment of the system, e.g., based on feedback of users. Our results show that the system performance, in terms of top-n accuracy and the mean reciprocal rank, benefits from re-ranking using gradient boosted regression trees. On average, the mean reciprocal rank improves by 9.15%.

cs.LG↗

arXiv2018

A categorisation and implementation of digital pen features for behaviour characterisation

In this paper we provide a categorisation and implementation of digital ink features for behaviour characterisation. Based on four feature sets taken from literature, we provide a categorisation in different classes of syntactic and semantic features. We implemented a publicly available framework to calculate these features and show its deployment in the use case of analysing cognitive assessments performed using a digital pen.

cs.CV↗

arXiv2018

A Survey on Deep Learning Toolkits and Libraries for Intelligent User Interfaces

This paper provides an overview of prominent deep learning toolkits and, in particular, reports on recent publications that contributed open source software for implementing tasks that are common in intelligent user interfaces (IUI). We provide a scientific reference for researchers and software engineers who plan to utilise deep learning techniques within their IUI research and development projects.

cs.HC↗

arXiv2017

Fine-tuning deep CNN models on specific MS COCO categories

Fine-tuning of a deep convolutional neural network (CNN) is often desired. This paper provides an overview of our publicly available py-faster-rcnn-ft software library that can be used to fine-tune the VGG_CNN_M_1024 model on custom subsets of the Microsoft Common Objects in Context (MS COCO) dataset. For example, we improved the procedure so that the user does not have to look for suitable image files in the dataset by hand which can then be used in the demo program. Our implementation randomly selects images that contain at least one object of the categories on which the model is fine-tuned.

cs.CV↗