@unpublished{AntonioRufinoJuniorRivaSanseverinoGalloetal.2023, author = {Ant{\^o}nio Rufino J{\´u}nior, Carlos and Riva Sanseverino, Eleonora and Gallo, Pierluigi and Amaral, Murilo Machado and Koch, Daniel and Kotak, Yash and Diel, Sergej and Walter, Gero and Schweiger, Hans-Georg and Zanin, Hudson}, title = {A Comprehensive Review of EV Lithium-Ion Battery Degradation}, publisher = {Preprints}, address = {Basel}, doi = {https://doi.org/10.20944/preprints202306.0228.v2}, year = {2023}, abstract = {Lithium-ion batteries with improved energy densities have made understanding the Solid Electrolyte Interphase (SEI) generation mechanisms that cause mechanical, thermal, and chemical failures more complicated. SEI processes reduce battery capacity and power. Thus, a review of this area's understanding is important. It is essential to know how batteries degrade in EVs to estimate battery lifespan as it goes, predict, and minimize losses, and determine the ideal time for a replacement. Lithium-ion batteries used in EVs mainly suffer two types of degradation: calendar degradation and cycling degradation. Despite the existence of several existing works in the literature, several aspects of battery degradation remain unclear or have not been analyzed in detail. This work presents a systematic review of existing works in the literature. The results of the present investigation provide insight into the complex relationships among various factors affecting battery degradation mechanisms. Specifically, this systematic review examined the effects of time, side reactions, temperature fluctuations, high charge/discharge rates, depth of discharge, mechanical stress, thermal stress, and the voltage relationship on battery performance and longevity. The results revealed that these factors interact in complex ways to influence the degradation mechanisms of batteries. For example, high charge currents and deep discharges were found to accelerate degradation, while low temperatures and moderate discharge depths were shown to be beneficial for battery longevity. Additionally, the results showed that the relationship between cell voltage and State-of-Charge (SOC) plays a critical role in determining the rate of degradation. Overall, these findings have important implications for the design and operation of battery systems, as they highlight the need to carefully manage a range of factors to maximize battery performance and longevity. The result is an analysis of the main articles published in this field in recent years. This work aims to present new knowledge about fault detection, diagnosis, and management of lithium-ion batteries based on battery degradation concepts. The new knowledge is presented and discussed in a structured and comprehensive way.}, language = {en} } @unpublished{FlohrSchussWallachetal.2023, author = {Flohr, Lukas A. and Schuß, Martina and Wallach, Dieter P. and Kr{\"u}ger, Antonio and Riener, Andreas}, title = {Designing for Passengers' Information Needs on Fellow Travelers: A Comparison of Day and Night Rides in Shared Automated Vehicles}, publisher = {arXiv}, address = {Ithaca}, doi = {https://doi.org/10.48550/arXiv.2308.02616}, year = {2023}, abstract = {Shared automated mobility-on-demand promises efficient, sustainable, and flexible transportation. Nevertheless, security concerns, resilience, and their mutual influence - especially at night - will likely be the most critical barriers to public adoption since passengers have to share rides with strangers without a human driver on board. As related work points out that information about fellow travelers might mitigate passengers' concerns, we designed two user interface variants to investigate the role of this information in an exploratory within-subjects user study (N = 24). Participants experienced four automated day and night rides with varying personal information about co-passengers in a simulated environment. The results of the mixed-method study indicate that having information about other passengers (e.g., photo, gender, and name) positively affects user experience at night. In contrast, it is less necessary during the day. Considering participants' simultaneously raised privacy demands poses a substantial challenge for resilient system design.}, language = {en} } @unpublished{SongLiuChenetal.2022, author = {Song, Rui and Liu, Dai and Chen, Dave Zhenyu and Festag, Andreas and Trinitis, Carsten and Schulz, Martin and Knoll, Alois}, title = {Federated Learning via Decentralized Dataset Distillation in Resource Constrained Edge Environments}, publisher = {arXiv}, address = {Ithaca}, doi = {https://doi.org/10.48550/arXiv.2208.11311}, year = {2022}, abstract = {In federated learning, all networked clients contribute to the model training cooperatively. However, with model sizes increasing, even sharing the trained partial models often leads to severe communication bottlenecks in underlying networks, especially when communicated iteratively. In this paper, we introduce a federated learning framework FedD3 requiring only one-shot communication by integrating dataset distillation instances. Instead of sharing model updates in other federated learning approaches, FedD3 allows the connected clients to distill the local datasets independently, and then aggregates those decentralized distilled datasets (e.g. a few unrecognizable images) from networks for model training. Our experimental results show that FedD3 significantly outperforms other federated learning frameworks in terms of needed communication volumes, while it provides the additional benefit to be able to balance the trade-off between accuracy and communication cost, depending on usage scenario or target dataset. For instance, for training an AlexNet model on CIFAR-10 with 10 clients under non-independent and identically distributed (Non-IID) setting, FedD3 can either increase the accuracy by over 71\% with a similar communication volume, or save 98\% of communication volume, while reaching the same accuracy, compared to other one-shot federated learning approaches.}, language = {en} } @unpublished{NovickiNetoRewayPolednaetal.2023, author = {Novicki Neto, Leonardo and Reway, Fabio and Poledna, Yuri and Funk Drechsler, Maikol and Parente Ribeiro, Eduardo and Huber, Werner and Icking, Christian}, title = {TWICE Dataset: Digital Twin of Test Scenarios in a Controlled Environment}, publisher = {arXiv}, address = {Ithaca}, doi = {https://doi.org/10.48550/arXiv.2310.03895}, year = {2023}, abstract = {Ensuring the safe and reliable operation of autonomous vehicles under adverse weather remains a significant challenge. To address this, we have developed a comprehensive dataset composed of sensor data acquired in a real test track and reproduced in the laboratory for the same test scenarios. The provided dataset includes camera, radar, LiDAR, inertial measurement unit (IMU), and GPS data recorded under adverse weather conditions (rainy, night-time, and snowy conditions). We recorded test scenarios using objects of interest such as car, cyclist, truck and pedestrian -- some of which are inspired by EURONCAP (European New Car Assessment Programme). The sensor data generated in the laboratory is acquired by the execution of simulation-based tests in hardware-in-the-loop environment with the digital twin of each real test scenario. The dataset contains more than 2 hours of recording, which totals more than 280GB of data. Therefore, it is a valuable resource for researchers in the field of autonomous vehicles to test and improve their algorithms in adverse weather conditions, as well as explore the simulation-to-reality gap.}, language = {en} } @unpublished{FidelisRewayRibeiroetal.2023, author = {Fidelis, Eduardo and Reway, Fabio and Ribeiro, Herick Y. S. and Campos, Pietro and Huber, Werner and Icking, Christian and Faria, Lester and Sch{\"o}n, Torsten}, title = {Generation of Realistic Synthetic Raw Radar Data for Automated Driving Applications using Generative Adversarial Networks}, publisher = {arXiv}, address = {Ithaca}, doi = {https://doi.org/10.48550/arXiv.2308.02632}, year = {2023}, abstract = {The main approaches for simulating FMCW radar are based on ray tracing, which is usually computationally intensive and do not account for background noise. This work proposes a faster method for FMCW radar simulation capable of generating synthetic raw radar data using generative adversarial networks (GAN). The code and pre-trained weights are open-source and available on GitHub. This method generates 16 simultaneous chirps, which allows the generated data to be used for the further development of algorithms for processing radar data (filtering and clustering). This can increase the potential for data augmentation, e.g., by generating data in non-existent or safety-critical scenarios that are not reproducible in real life. In this work, the GAN was trained with radar measurements of a motorcycle and used to generate synthetic raw radar data of a motorcycle traveling in a straight line. For generating this data, the distance of the motorcycle and Gaussian noise are used as input to the neural network. The synthetic generated radar chirps were evaluated using the Frechet Inception Distance (FID). Then, the Range-Azimuth (RA) map is calculated twice: first, based on synthetic data using this GAN and, second, based on real data. Based on these RA maps, an algorithm with adaptive threshold and edge detection is used for object detection. The results have shown that the data is realistic in terms of coherent radar reflections of the motorcycle and background noise based on the comparison of chirps, the RA maps and the object detection results. Thus, the proposed method in this work has shown to minimize the simulation-to-reality gap for the generation of radar data.}, language = {en} } @unpublished{HanKefferpuetzBeyerer2025, author = {Han, Longfei and Kefferp{\"u}tz, Klaus and Beyerer, J{\"u}rgen}, title = {Decentralized Fusion of 3D Extended Object Tracking based on a B-Spline Shape Model}, publisher = {arXiv}, address = {Ithaca}, doi = {https://doi.org/10.48550/arXiv.2504.18708}, year = {2025}, abstract = {Extended Object Tracking (EOT) exploits the high resolution of modern sensors for detailed environmental perception. Combined with decentralized fusion, it contributes to a more scalable and robust perception system. This paper investigates the decentralized fusion of 3D EOT using a B-spline curve based model. The spline curve is used to represent the side-view profile, which is then extruded with a width to form a 3D shape. We use covariance intersection (CI) for the decentralized fusion and discuss the challenge of applying it to EOT. We further evaluate the tracking result of the decentralized fusion with simulated and real datasets of traffic scenarios. We show that the CI-based fusion can significantly improve the tracking performance for sensors with unfavorable perspective.}, language = {en} } @unpublished{KruegelOstermaierUhl2023, author = {Kr{\"u}gel, Sebastian and Ostermaier, Andreas and Uhl, Matthias}, title = {The moral authority of ChatGPT}, publisher = {arXiv}, address = {Ithaca}, doi = {https://doi.org/10.48550/arXiv.2301.07098}, year = {2023}, abstract = {ChatGPT is not only fun to chat with, but it also searches information, answers questions, and gives advice. With consistent moral advice, it might improve the moral judgment and decisions of users, who often hold contradictory moral beliefs. Unfortunately, ChatGPT turns out highly inconsistent as a moral advisor. Nonetheless, it influences users' moral judgment, we find in an experiment, even if they know they are advised by a chatting bot, and they underestimate how much they are influenced. Thus, ChatGPT threatens to corrupt rather than improves users' judgment. These findings raise the question of how to ensure the responsible use of ChatGPT and similar AI. Transparency is often touted but seems ineffective. We propose training to improve digital literacy.}, language = {en} } @unpublished{KammerlanderKolbLuegmairetal.2025, author = {Kammerlander, Calvin and Kolb, Viola and Luegmair, Marinus and Scheermann, Lou and Schmailzl, Maximilian and Seufert, Marco and Zhang, Jiayun and Dalic, Denis and Sch{\"o}n, Torsten}, title = {Machine Learning Models for Soil Parameter Prediction Based on Satellite, Weather, Clay and Yield Data}, publisher = {arXiv}, address = {Ithaca}, doi = {https://doi.org/10.48550/arXiv.2503.22276}, year = {2025}, abstract = {Efficient nutrient management and precise fertilization are essential for advancing modern agriculture, particularly in regions striving to optimize crop yields sustainably. The AgroLens project endeavors to address this challenge by develop ing Machine Learning (ML)-based methodologies to predict soil nutrient levels without reliance on laboratory tests. By leveraging state of the art techniques, the project lays a foundation for acionable insights to improve agricultural productivity in resource-constrained areas, such as Africa. The approach begins with the development of a robust European model using the LUCAS Soil dataset and Sentinel-2 satellite imagery to estimate key soil properties, including phosphorus, potassium, nitrogen, and pH levels. This model is then enhanced by integrating supplementary features, such as weather data, harvest rates, and Clay AI-generated embeddings. This report details the methodological framework, data preprocessing strategies, and ML pipelines employed in this project. Advanced algorithms, including Random Forests, Extreme Gradient Boosting (XGBoost), and Fully Connected Neural Networks (FCNN), were implemented and finetuned for precise nutrient prediction. Results showcase robust model performance, with root mean square error values meeting stringent accuracy thresholds. By establishing a reproducible and scalable pipeline for soil nutrient prediction, this research paves the way for transformative agricultural applications, including precision fertilization and improved resource allocation in underresourced regions like Africa.}, language = {en} } @unpublished{AxenieSaveriano2023, author = {Axenie, Cristian and Saveriano, Matteo}, title = {Antifragile Control Systems: The case of mobile robot trajectory tracking in the presence of uncertainty}, publisher = {arXiv}, address = {Ithaca}, doi = {https://doi.org/10.48550/arXiv.2302.05117}, year = {2023}, abstract = {Mobile robots are ubiquitous. Such vehicles benefit from well-designed and calibrated control algorithms ensuring their task execution under precise uncertainty bounds. Yet, in tasks involving humans in the loop, such as elderly or mobility impaired, the problem takes a new dimension. In such cases, the system needs not only to compensate for uncertainty and volatility in its operation but at the same time to anticipate and offer responses that go beyond robust. Such robots operate in cluttered, complex environments, akin to human residences, and need to face during their operation sensor and, even, actuator faults, and still operate. This is where our thesis comes into the foreground. We propose a new control design framework based on the principles of antifragility. Such a design is meant to offer a high uncertainty anticipation given previous exposure to failures and faults, and exploit this anticipation capacity to provide performance beyond robust. In the current instantiation of antifragile control applied to mobile robot trajectory tracking, we provide controller design steps, the analysis of performance under parametrizable uncertainty and faults, as well as an extended comparative evaluation against state-of-the-art controllers. We believe in the potential antifragile control has in achieving closed-loop performance in the face of uncertainty and volatility by using its exposures to uncertainty to increase its capacity to anticipate and compensate for such events.}, language = {en} } @unpublished{SongLiangXiaetal.2025, author = {Song, Rui and Liang, Chenwei and Xia, Yan and Zimmer, Walter and Cao, Hu and Caesar, Holger and Festag, Andreas and Knoll, Alois}, title = {CoDa-4DGS: Dynamic Gaussian Splatting with Context and Deformation Awareness for Autonomous Driving}, publisher = {arXiv}, address = {Ithaca}, doi = {https://doi.org/10.48550/arXiv.2503.06744}, year = {2025}, abstract = {Dynamic scene rendering opens new avenues in autonomous driving by enabling closed-loop simulations with photorealistic data, which is crucial for validating end-to-end algorithms. However, the complex and highly dynamic nature of traffic environments presents significant challenges in accurately rendering these scenes. In this paper, we introduce a novel 4D Gaussian Splatting (4DGS) approach, which incorporates context and temporal deformation awareness to improve dynamic scene rendering. Specifically, we employ a 2D semantic segmentation foundation model to self-supervise the 4D semantic features of Gaussians, ensuring meaningful contextual embedding. Simultaneously, we track the temporal deformation of each Gaussian across adjacent frames. By aggregating and encoding both semantic and temporal deformation features, each Gaussian is equipped with cues for potential deformation compensation within 3D space, facilitating a more precise representation of dynamic scenes. Experimental results show that our method improves 4DGS's ability to capture fine details in dynamic scene rendering for autonomous driving and outperforms other self-supervised methods in 4D reconstruction and novel view synthesis. Furthermore, CoDa-4DGS deforms semantic features with each Gaussian, enabling broader applications.}, language = {en} } @unpublished{HeinlPatapovasPilgermann2024, author = {Heinl, Patrizia and Patapovas, Andrius and Pilgermann, Michael}, title = {Towards AI-enabled Cyber Threat Assessment in the Health Sector}, publisher = {arXiv}, address = {Ithaca}, doi = {https://doi.org/10.48550/arXiv.2409.12765}, year = {2024}, abstract = {Cyber attacks on the healthcare industry can have tremendous consequences and the attack surface expands continuously. In order to handle the steadily rising workload, an expanding amount of analog processes in healthcare institutions is digitized. Despite regulations becoming stricter, not all existing infrastructure is sufficiently protected against cyber attacks. With an increasing number of devices and digital processes, the system and network landscape becomes more complex and harder to manage and therefore also more difficult to protect. The aim of this project is to introduce an AI-enabled platform that collects security relevant information from the outside of a health organization, analyzes it, delivers a risk score and supports decision makers in healthcare institutions to optimize investment choices for security measures. Therefore, an architecture of such a platform is designed, relevant information sources are identified, and AI methods for relevant data collection, selection, and risk scoring are explored.}, language = {en} } @unpublished{KalyanasundaramChandraSekaranStaeuberetal.2025, author = {Kalyanasundaram, Abinav and Chandra Sekaran, Karthikeyan and St{\"a}uber, Philipp and Lange, Michael and Utschick, Wolfgang and Botsch, Michael}, title = {Uncertainty-Aware Hybrid Machine Learning in Virtual Sensors for Vehicle Sideslip Angle Estimation}, publisher = {arXiv}, address = {Ithaca}, doi = {https://doi.org/10.48550/arXiv.2504.06105}, year = {2025}, abstract = {Precise vehicle state estimation is crucial for safe and reliable autonomous driving. The number of measurable states and their precision offered by the onboard vehicle sensor system are often constrained by cost. For instance, measuring critical quantities such as the Vehicle Sideslip Angle (VSA) poses significant commercial challenges using current optical sensors. This paper addresses these limitations by focusing on the development of high-performance virtual sensors to enhance vehicle state estimation for active safety. The proposed Uncertainty-Aware Hybrid Learning (UAHL) architecture integrates a machine learning model with vehicle motion models to estimate VSA directly from onboard sensor data. A key aspect of the UAHL architecture is its focus on uncertainty quantification for individual model estimates and hybrid fusion. These mechanisms enable the dynamic weighting of uncertainty-aware predictions from machine learning and vehicle motion models to produce accurate and reliable hybrid VSA estimates. This work also presents a novel dataset named Real-world Vehicle State Estimation Dataset (ReV-StED), comprising synchronized measurements from advanced vehicle dynamic sensors. The experimental results demonstrate the superior performance of the proposed method for VSA estimation, highlighting UAHL as a promising architecture for advancing virtual sensors and enhancing active safety in autonomous vehicles.}, language = {en} } @unpublished{HanXuKefferpuetzetal.2024, author = {Han, Longfei and Xu, Qiuyu and Kefferp{\"u}tz, Klaus and Elger, Gordon and Beyerer, J{\"u}rgen}, title = {Applying Extended Object Tracking for Self-Localization of Roadside Radar Sensors}, publisher = {arXiv}, address = {Ithaca}, doi = {https://doi.org/10.48550/arXiv.2407.03084}, year = {2024}, abstract = {Intelligent Transportation Systems (ITS) can benefit from roadside 4D mmWave radar sensors for large-scale traffic monitoring due to their weatherproof functionality, long sensing range and low manufacturing cost. However, the localization method using external measurement devices has limitations in urban environments. Furthermore, if the sensor mount exhibits changes due to environmental influences, they cannot be corrected when the measurement is performed only during the installation. In this paper, we propose self-localization of roadside radar data using Extended Object Tracking (EOT). The method analyses both the tracked trajectories of the vehicles observed by the sensor and the aerial laser scan of city streets, assigns labels of driving behaviors such as "straight ahead", "left turn", "right turn" to trajectory sections and road segments, and performs Semantic Iterative Closest Points (SICP) algorithm to register the point cloud. The method exploits the result from a down stream task -- object tracking -- for localization. We demonstrate high accuracy in the sub-meter range along with very low orientation error. The method also shows good data efficiency. The evaluation is done in both simulation and real-world tests.}, language = {en} } @unpublished{HanKefferpuetzBeyerer2025, author = {Han, Longfei and Kefferp{\"u}tz, Klaus and Beyerer, J{\"u}rgen}, title = {3D Extended Object Tracking based on Extruded B-Spline Side View Profiles}, publisher = {arXiv}, address = {Ithaca}, doi = {https://doi.org/10.48550/arXiv.2503.10730}, year = {2025}, abstract = {Object tracking is an essential task for autonomous systems. With the advancement of 3D sensors, these systems can better perceive their surroundings using effective 3D Extended Object Tracking (EOT) methods. Based on the observation that common road users are symmetrical on the right and left sides in the traveling direction, we focus on the side view profile of the object. In order to leverage of the development in 2D EOT and balance the number of parameters of a shape model in the tracking algorithms, we propose a method for 3D extended object tracking (EOT) by describing the side view profile of the object with B-spline curves and forming an extrusion to obtain a 3D extent. The use of B-spline curves exploits their flexible representation power by allowing the control points to move freely. The algorithm is developed into an Extended Kalman Filter (EKF). For a through evaluation of this method, we use simulated traffic scenario of different vehicle models and realworld open dataset containing both radar and lidar data.}, language = {en} } @unpublished{PaulaBauderPfeilschifteretal.2023, author = {Paula, Daniel and Bauder, Maximilian and Pfeilschifter, Claus and Petermeier, Franziska and Kubjatko, Tibor and B{\"o}hm, Klaus and Riener, Andreas and Schweiger, Hans-Georg}, title = {Impact of Partially Automated Driving Functions on Forensic Accident Reconstruction: A Simulator Study on Driver Reaction Behavior in the Event of a Malfunctioning System Behavior}, publisher = {Preprints}, address = {Basel}, doi = {https://doi.org/10.20944/preprints202311.0947.v1}, year = {2023}, abstract = {Partially automated driving functions (SAE Level 2) can control a vehicle's longitudinal and lateral movements. However, taking over the driving task involves automation risks that the driver must manage. In severe accidents, the driver's ability to avoid a collision must be assessed, considering their expected reaction behavior. The primary goal of this study is to generate essential data on driver reaction behavior in case of malfunctions in partially automated driving functions for use in legal affairs. A simulator study with two scenarios involving 32 subjects was conducted for this purpose. The first scenario investigated driver reactions to system limitations during cornering. The second scenario examined driver responses to phantom braking caused by the AEBS. As a result, the first scenario shows that none of the subjects could control the situation safely. Due to partial automation, we could also identify a new part of the reaction time, the hands-on time, which leads to increased steering reaction times of 1.18 to 1.74 seconds. In the second scenario, we found that 25 of the 32 subjects could not override the phantom braking by pressing the accelerator pedal, although 16 subjects were informed about the system analog to the actual vehicle manuals. Overall, the study suggests that the current legal perspective on vehicle control and the expected driver reaction behavior for accident avoidance should be reconsidered.}, language = {en} } @unpublished{ChaturvediKumarLambaetal.2024, author = {Chaturvedi, Pushpendra Kr. and Kumar, Nand and Lamba, Ravita and Mehta, Kedar}, title = {Multi-objective optimization of glazing and shading configurations for visual, thermal, and energy performance of cooling dominant climatic regions of India}, titleParent = {Research Square}, publisher = {Research Square}, address = {Durham}, doi = {https://doi.org/10.21203/rs.3.rs-4983831/v1}, year = {2024}, abstract = {Climate adaptive passive design features such as glazing and shading configurations, often exhibit conflicting behaviour to maintaining indoor visual and thermal environment while minimizing energy consumption. This study employed the MOO approach through SPEA-II and HypE algorithms to find the efficient glazing and shading configurations in four cooling dominant climatic zones of India such as hot and dry, composite, warm and humid, and temperate. A residential building situated in Jaipur city (India) was chosen for the analysis and to demonstrate the effectiveness and reliability of the optimization process. Twenty-six design variables including wall window ratio, louvers depth, louvers count, fin depth and fins count of each orientation, window height, sill height, glazing U-value, solar heat gain coefficient, and visual light transmittance were imported into the Octopus to investigate their interactive impact on the Useful Daylight Illuminance (UDI), Thermal Discomfort Percentage (TDP) and Energy Use Intensity (EUI). The optimized trade-off solutions represented a considerable improvement; UDI increased by 24.61\%, 21.90\%, 14.91\% \& 26.41\%, and TDP reduced by 10.38\%, 1.5\%, 13.95\%, \& 17.68\%, and EUI decreased by 10.10\%, 1.9\%, 15.82\%, \& 19.19\% compared to initially generated solutions for Ahmedabad, Delhi, Mumbai and Bangalore cities respectively. Finally, the Bayesian machine learning technique has been used for sensitivity analysis to identify the correlation between the design variables and performance objectives. The outcomes of this research can help architects in selecting optimal glazing and shading features for both new building design and retrofitting of existing buildings in the tropical climatic conditions of India.}, language = {en} } @unpublished{EndtEngelNaldietal.2023, author = {Endt, Sebastian and Engel, Maria and Naldi, Emanuele and Assereto, Rodolfo and Molendowska, Malwina and Mueller, Lars and Verdun, Claudio Mayrink and Pirkl, Carolin and Palombo, Marco and Jones, Derek K. and Menzel, Marion Irene}, title = {In-vivo myelin water quantification using diffusion-relaxation correlation MRI: a comparison of 1D and 2D methods}, titleParent = {Research Square}, publisher = {Research Square}, address = {Durham}, doi = {https://doi.org/10.21203/rs.3.rs-3069146/v1}, year = {2023}, abstract = {Multidimensional Magnetic Resonance Imaging (MRI) is a versatile tool for microstructure mapping. We use a diffusion weighted inversion-recovery spin echo (DW-IR-SE) sequence with spiral readouts at ultra-strong gradients to acquire a rich diffusion-relaxation data set with sensitivity to myelin water. We reconstruct 1D and 2D spectra with a two-step convex optimization approach and investigate a variety of multidimensional MRI methods, including 1D multi-component relaxometry, 1D multi-component diffusometry, 2D relaxation correlation imaging, and 2D diffusion-relaxation correlation spectroscopic imaging (DR-CSI), in terms of their potential to quantify tissue microstructure, including the myelin water fraction (MWF). We observe a distinct spectral peak that we attribute to myelin water in multi-component T1 relaxometry, T1-T2 correlation, T1-D correlation, and T2-D correlation imaging. Due to lower achievable echo times compared to diffusometry, MWF maps from relaxometry have higher quality. While 1D multi-component T1 data allows much faster myelin mapping, 2D approaches could offer unique insights into tissue microstructure and especially myelin diffusion.}, language = {en} } @unpublished{KruegelUhl2022, author = {Kr{\"u}gel, Sebastian and Uhl, Matthias}, title = {The risk ethics of autonomous vehicles: a continuous trolley problem in regular road traffic}, publisher = {arXiv}, address = {Ithaca}, doi = {https://doi.org/10.48550/arXiv.2206.03258}, year = {2022}, abstract = {Is the ethics of autonomous vehicles (AVs) restricted to weighing lives in unavoidable accidents? We argue that AVs distribute risks between road users in regular traffic situations, either explicitly or implicitly. This distribution of risks raises ethically relevant questions that cannot be evaded by simple heuristics such as "hitting the brakes." Using an interactive, graphical representation of different traffic situations, we measured participants' preferences on driving maneuvers of AVs in a representative survey in Germany. Our participants' preferences deviated significantly from mere collision avoidance. Interestingly, our participants were willing to take risks themselves for the benefit of other road users suggesting that the social dilemma of AVs may lessen in a context of risk.}, language = {en} } @unpublished{PetrovskaErjiageKugele2025, author = {Petrovska, Ana and Erjiage, Guan and Kugele, Stefan}, title = {Defining Self-adaptive Systems: A Systematic Literature Review}, publisher = {arXiv}, address = {Ithaca}, doi = {https://doi.org/10.48550/arXiv.2505.17798}, year = {2025}, abstract = {In the last two decades, the popularity of self-adaptive systems in the field of software and systems engineering has drastically increased. However, despite the extensive work on self-adaptive systems, the literature still lacks a common agreement on the definition of these systems. To this day, the notion of self-adaptive systems is mainly used intuitively without a precise understanding of the terminology. Using terminology only by intuition does not suffice, especially in engineering and science, where a more rigorous definition is necessary. In this paper, we investigate the existing formal definitions of self-adaptive systems and how these systems are characterised across the literature. Additionally, we analyse and summarise the limitations of the existing formal definitions in order to understand why none of the existing formal definitions is used more broadly by the community. To achieve this, we have conducted a systematic literature review in which we have analysed over 1400 papers related to self-adaptive systems. Concretely, from an initial pool of 1493 papers, we have selected 314 relevant papers, which resulted in nine primary studies whose primary objective was to define self-adaptive systems formally. Our systematic review reveals that although there has been an increasing interest in self-adaptive systems over the years, there is a scarcity of efforts to define these systems formally. Finally, as part of this paper, based on the analysed primary studies, we also elicit requirements and set a foundation for a potential (formal) definition in the future that is accepted by the community on a broader range.}, language = {en} } @unpublished{RamicKugele2025, author = {Ramic, Amra and Kugele, Stefan}, title = {A Systematic Mapping Study on Software Architecture for AI-based Mobility Systems}, publisher = {arXiv}, address = {Ithaca}, doi = {https://doi.org/10.48550/arXiv.2506.01595}, year = {2025}, abstract = {Background: Due to their diversity, complexity, and above all importance, safety-critical and dependable systems must be developed with special diligence. Criticality increases as these systems likely contain artificial intelligence (AI) components known for their uncertainty. As software and reference architectures form the backbone of any successful system, including safety-critical dependable systems with learning-enabled components, choosing the suitable architecture that guarantees safety despite uncertainties is of great eminence. Aim: We aim to provide the missing overview of all existing architectures, their contribution to safety, and their level of maturity in AI-based safety-critical systems. Method: To achieve this aim, we report a systematic mapping study. From a set of 1,639 primary studies, we selected 38 relevant studies dealing with safety assurance through software architecture in AI-based safety-critical systems. The selected studies were then examined using various criteria to answer the research questions and identify gaps in this area of research. Results: Our findings showed which architectures have been proposed and to what extent they have been implemented. Furthermore, we identified gaps in different application areas of those systems and explained these gaps with various arguments. Conclusion: As the AI trend continues to grow, the system complexity will inevitably increase, too. To ensure the lasting safety of the systems, we provide an overview of the state of the art, intending to identify best practices and research gaps and direct future research more focused.}, language = {en} } @unpublished{PederivaDeMartinoZimmer2023, author = {Pederiva, Marcelo Eduardo and De Martino, Jos{\´e} Mario and Zimmer, Alessandro}, title = {MonoNext: A 3D Monocular Object Detection with ConvNext}, publisher = {arXiv}, address = {Ithaca}, doi = {https://doi.org/10.48550/arXiv.2308.00596}, year = {2023}, abstract = {Autonomous driving perception tasks rely heavily on cameras as the primary sensor for Object Detection, Semantic Segmentation, Instance Segmentation, and Object Tracking. However, RGB images captured by cameras lack depth information, which poses a significant challenge in 3D detection tasks. To supplement this missing data, mapping sensors such as LIDAR and RADAR are used for accurate 3D Object Detection. Despite their significant accuracy, the multi-sensor models are expensive and require a high computational demand. In contrast, Monocular 3D Object Detection models are becoming increasingly popular, offering a faster, cheaper, and easier-to-implement solution for 3D detections. This paper introduces a different Multi-Tasking Learning approach called MonoNext that utilizes a spatial grid to map objects in the scene. MonoNext employs a straightforward approach based on the ConvNext network and requires only 3D bounding box annotated data. In our experiments with the KITTI dataset, MonoNext achieved high precision and competitive performance comparable with state-of-the-art approaches. Furthermore, by adding more training data, MonoNext surpassed itself and achieved higher accuracies.}, language = {en} } @unpublished{GleirscherLinkerKugele2018, author = {Gleirscher, Mario and Linker, Sven and Kugele, Stefan}, title = {SCAV'18: Report of the 2nd International Workshop on Safe Control of Autonomous Vehicles}, publisher = {arXiv}, address = {Ithaca}, doi = {https://doi.org/10.48550/arXiv.1811.01774}, year = {2018}, abstract = {This report summarizes the discussions, open issues, take-away messages, and conclusions of the 2nd SCAV workshop.}, language = {en} } @unpublished{BalajiBliesGoerietal.2021, author = {Balaji, Thangapavithraa and Blies, Patrick and G{\"o}ri, Georg and Mitsch, Raphael and Wasserer, Marcel and Sch{\"o}n, Torsten}, title = {Temporally coherent video anonymization through GAN inpainting}, publisher = {arXiv}, address = {Ithaca}, doi = {https://doi.org/10.48550/arXiv.2106.02328}, year = {2021}, language = {en} } @unpublished{BazziSepulcreDeloozetal.2023, author = {Bazzi, Alessandro and Sepulcre, Miguel and Delooz, Quentin and Festag, Andreas and Vogt, Jonas and Wieker, Horst and Berens, Friedbert and Spaanderman, Paul}, title = {Multi-Channel Operation for the Release 2 of ETSI Cooperative Intelligent Transport Systems}, publisher = {arXiv}, address = {Ithaca}, doi = {https://doi.org/10.48550/arXiv.2305.19863}, year = {2023}, abstract = {Vehicles and road infrastructure are starting to be equipped with vehicle-to-everything (V2X) communication solutions to increase road safety and provide new services to drivers and passengers. In Europe, the deployment is based on a set of Release 1 standards developed by ETSI to support basic use cases for cooperative intelligent transport systems (C-ITS). For them, the capacity of a single 10 MHz channel in the ITS band at 5.9 GHz is considered sufficient. At the same time, the ITS stakeholders are working towards several advanced use cases, which imply a significant increment of data traffic and the need for multiple channels. To address this issue, ETSI has recently standardized a new multi-channel operation (MCO) concept for flexible, efficient, and future-proof use of multiple channels. This new concept is defined in a set of new specifications that represent the foundation for the future releases of C-ITS standards. The present paper provides a comprehensive review of the new set of specifications, describing the main entities extending the C-ITS architecture at the different layers of the protocol stack, In addition, the paper provides representative examples that describe how these MCO standards will be used in the future and discusses some of the main open issues arising. The review and analysis of this paper facilitate the understanding and motivation of the new set of Release 2 ETSI specifications for MCO and the identification of new research opportunities.}, language = {en} } @unpublished{SongLiangCaoetal.2024, author = {Song, Rui and Liang, Chenwei and Cao, Hu and Yan, Zhiran and Zimmer, Walter and Gross, Markus and Festag, Andreas and Knoll, Alois}, title = {Collaborative Semantic Occupancy Prediction with Hybrid Feature Fusion in Connected Automated Vehicles}, publisher = {arXiv}, address = {Ithaca}, doi = {https://doi.org/10.48550/arXiv.2402.07635}, year = {2024}, abstract = {Collaborative perception in automated vehicles leverages the exchange of information between agents, aiming to elevate perception results. Previous camera-based collaborative 3D perception methods typically employ 3D bounding boxes or bird's eye views as representations of the environment. However, these approaches fall short in offering a comprehensive 3D environmental prediction. To bridge this gap, we introduce the first method for collaborative 3D semantic occupancy prediction. Particularly, it improves local 3D semantic occupancy predictions by hybrid fusion of (i) semantic and occupancy task features, and (ii) compressed orthogonal attention features shared between vehicles. Additionally, due to the lack of a collaborative perception dataset designed for semantic occupancy prediction, we augment a current collaborative perception dataset to include 3D collaborative semantic occupancy labels for a more robust evaluation. The experimental findings highlight that: (i) our collaborative semantic occupancy predictions excel above the results from single vehicles by over 30\%, and (ii) models anchored on semantic occupancy outpace state-of-the-art collaborative 3D detection techniques in subsequent perception applications, showcasing enhanced accuracy and enriched semantic-awareness in road environments.}, language = {en} } @unpublished{PhanSommerRitzetal.2022, author = {Phan, Thomy and Sommer, Felix and Ritz, Fabian and Altmann, Philipp and N{\"u}ßlein, Jonas and K{\"o}lle, Michael and Belzner, Lenz and Linnhoff-Popien, Claudia}, title = {Emergent Cooperation from Mutual Acknowledgment Exchange in Multi-Agent Reinforcement Learning}, titleParent = {Research Square}, publisher = {Research Square}, address = {Durham}, doi = {https://doi.org/10.21203/rs.3.rs-2315844/v1}, year = {2022}, abstract = {Peer incentivization (PI) is a recent approach, where all agents learn to reward or to penalize each other in a distributed fashion which often leads to emergent cooperation. Current PI mechanisms implicitly assume a flawless communication channel in order to exchange rewards. These rewards are directly integrated into the learning process without any chance to respond with feedback. Furthermore, most PI approaches rely on global information which limits scalability and applicability to real-world scenarios, where only local information is accessible. In this paper, we propose Mutual Acknowledgment Token Exchange (MATE), a PI approach defined by a two-phase communication protocol to mutually exchange acknowledgment tokens to shape individual rewards. Each agent evaluates the monotonic improvement of its individual situation in order to accept or reject acknowledgment requests from other agents. MATE is completely decentralized and only requires local communication and information. We evaluate MATE in three social dilemma domains. Our results show that MATE is able to achieve and maintain significantly higher levels of cooperation than previous PI approaches. In addition, we evaluate the robustness of MATE in more realistic scenarios, where agents can defect from the protocol and where communication failures can occur. We also evaluate the sensitivity of MATE w.r.t. the choice of token values.}, language = {en} } @unpublished{RoessleGernerBogenbergeretal.2024, author = {R{\"o}ßle, Dominik and Gerner, Jeremias and Bogenberger, Klaus and Cremers, Daniel and Schmidtner, Stefanie and Sch{\"o}n, Torsten}, title = {Unlocking Past Information: Temporal Embeddings in Cooperative Bird's Eye View Prediction}, publisher = {arXiv}, address = {Ithaca}, doi = {https://doi.org/10.48550/arXiv.2401.14325}, year = {2024}, abstract = {Accurate and comprehensive semantic segmentation of Bird's Eye View (BEV) is essential for ensuring safe and proactive navigation in autonomous driving. Although cooperative perception has exceeded the detection capabilities of single-agent systems, prevalent camera-based algorithms in cooperative perception neglect valuable information derived from historical observations. This limitation becomes critical during sensor failures or communication issues as cooperative perception reverts to single-agent perception, leading to degraded performance and incomplete BEV segmentation maps. This paper introduces TempCoBEV, a temporal module designed to incorporate historical cues into current observations, thereby improving the quality and reliability of BEV map segmentations. We propose an importance-guided attention architecture to effectively integrate temporal information that prioritizes relevant properties for BEV map segmentation. TempCoBEV is an independent temporal module that seamlessly integrates into state-of-the-art camera-based cooperative perception models. We demonstrate through extensive experiments on the OPV2V dataset that TempCoBEV performs better than non-temporal models in predicting current and future BEV map segmentations, particularly in scenarios involving communication failures. We show the efficacy of TempCoBEV and its capability to integrate historical cues into the current BEV map, improving predictions under optimal communication conditions by up to 2\% and under communication failures by up to 19\%. The code will be published on GitHub.}, language = {en} } @unpublished{HashemiKřetinskyRiederetal.2024, author = {Hashemi, Vahid and Křet{\´i}nsk{\´y}, Jan and Rieder, Sabine and Sch{\"o}n, Torsten and Vorhoff, Jan}, title = {Gaussian-Based and Outside-the-Box Runtime Monitoring Join Forces}, publisher = {arXiv}, address = {Ithaca}, doi = {https://doi.org/10.48550/arXiv.2410.06051}, year = {2024}, abstract = {Since neural networks can make wrong predictions even with high confidence, monitoring their behavior at runtime is important, especially in safety-critical domains like autonomous driving. In this paper, we combine ideas from previous monitoring approaches based on observing the activation values of hidden neurons. In particular, we combine the Gaussian-based approach, which observes whether the current value of each monitored neuron is similar to typical values observed during training, and the Outside-the-Box monitor, which creates clusters of the acceptable activation values, and, thus, considers the correlations of the neurons' values. Our experiments evaluate the achieved improvement.}, language = {en} } @unpublished{OgnawalaPretschnerHutzelmannetal.2018, author = {Ognawala, Saahil and Pretschner, Alexander and Hutzelmann, Thomas and Psallida, Eirini and Amato, Ricardo Nales}, title = {Reviewing KLEE's Sonar-Search Strategy in Context of Greybox Fuzzing}, publisher = {arXiv}, address = {Ithaca}, doi = {https://doi.org/10.48550/arXiv.1803.04881}, year = {2018}, abstract = {Automatic test-case generation techniques of symbolic execution and fuzzing are the most widely used methods to discover vulnerabilities in, both, academia and industry. However, both these methods suffer from fundamental drawbacks that stop them from achieving high path coverage that may, consequently, lead to discovering vulnerabilities at the numerical scale of static analysis. In this presentation, we examine systems-under-test (SUTs) at the granularity level of functions and postulate that achieving higher function coverage (execution of functions in a program at least once) than, both, symbolic execution and fuzzing may be a necessary condition for discovering more vulnerabilities than both. We will start this presentation with the design of a targeted search strategy for KLEE, sonar-search, that prioritizes paths leading to a target function, rather than maximizing overall path coverage in the program. Then, we will show that examining SUTs at the level of functions (compositional analysis) leads to discovering more vulnerabilities than symbolic execution from a single entry point. Using this finding, we will, then, demonstrate a greybox fuzzing method that can achieve higher function coverage than symbolic execution. Finally, we will present a framework to effectively manage vulnerabilities and assess their severities.}, language = {en} } @unpublished{StreicherBielefeldEller2022, author = {Streicher, Bernhard and Bielefeld, Moritz and Eller, Eric}, title = {The Risk Culture Framework: Introducing an integrative framework for holistic risk analysis}, publisher = {Advance}, address = {London}, doi = {https://doi.org/10.31124/advance.19104632.v1}, year = {2022}, language = {en} } @unpublished{NanDelSerWalshetal.2022, author = {Nan, Yang and Del Ser, Javier and Walsh, Simon and Sch{\"o}nlieb, Carola and Roberts, Michael and Selby, Ian and Howard, Kit and Owen, John and Neville, Jon and Guiot, Julien and Ernst, Benoit and Pastor, Ana and Alberich-Bayarri, Angel and Menzel, Marion Irene and Walsh, Sean and Vos, Wim and Flerin, Nina and Charbonnier, Jean-Paul and Rikxoort, Eva van and Chatterjee, Avishek and Woodruff, Henry and Lambin, Philippe and Cerd{\´a}-Alberich, Leonor and Mart{\´i}-Bonmat{\´i}, Luis and Herrera, Francisco and Yang, Guang}, title = {Data Harmonisation for Information Fusion in Digital Healthcare: A State-of-the-Art Systematic Review, Meta-Analysis and Future Research Directions}, publisher = {arXiv}, address = {Ithaca}, doi = {https://doi.org/10.48550/arXiv.2201.06505}, year = {2022}, abstract = {Removing the bias and variance of multicentre data has always been a challenge in large scale digital healthcare studies, which requires the ability to integrate clinical features extracted from data acquired by different scanners and protocols to improve stability and robustness. Previous studies have described various computational approaches to fuse single modality multicentre datasets. However, these surveys rarely focused on evaluation metrics and lacked a checklist for computational data harmonisation studies. In this systematic review, we summarise the computational data harmonisation approaches for multi-modality data in the digital healthcare field, including harmonisation strategies and evaluation metrics based on different theories. In addition, a comprehensive checklist that summarises common practices for data harmonisation studies is proposed to guide researchers to report their research findings more effectively. Last but not least, flowcharts presenting possible ways for methodology and metric selection are proposed and the limitations of different methods have been surveyed for future research.}, language = {en} } @unpublished{AfrajVaculinBoehmlaenderetal.2023, author = {Afraj, Shahabaz and Vaculin, Ondrej and B{\"o}hml{\"a}nder, Dennis and Hynč{\´i}k, Luděk}, title = {Vehicle Crash Simulation Models for Reinforcement Learning driven crash-detection algorithm calibration}, titleParent = {Research Square}, publisher = {Research Square}, address = {Durham}, doi = {https://doi.org/10.21203/rs.3.rs-3004299/v1}, year = {2023}, abstract = {The development of finite element vehicle models for crash simulations is a highly complex task. The main aim of these models is to simulate a variety of crash scenarios and assess all the safety systems for their respective performances. These vehicle models possess a substantial amount of data pertaining to the vehicle's geometry, structure, materials, etc., and are used to estimate a large set of system and component level characteristics using crash simulations. It is understood that even the most well-developed simulation models are prone to deviations in estimation when compared to real-world physical test results. This is generally due to our inability to model the chaos and uncertainties introduced in the real world. Such unavoidable deviations render the use of virtual simulations ineffective for the calibration process of the algorithms that activate the restraint systems in the event of a crash (crash-detection algorithm). In the scope of this research, authors hypothesize the possibility of accounting for such variations introduced in the real world by creating a feedback loop between real-world crash tests and crash simulations. To accomplish this, a Reinforcement Learning (RL) compatible virtual surrogate model is used, which is adapted from crash simulation models. Hence, a conceptual methodology is illustrated in this paper for developing an RL-compatible model that can be trained using the results of crash simulations and crash tests. As the calibration of the crash-detection algorithm is fundamentally dependent upon the crash pulses, the scope of the expected output is limited to advancing the ability to estimate crash pulses. Furthermore, the real-time implementation of the methodology is illustrated using an actual vehicle model.}, language = {en} } @unpublished{GlockerHerygersGeorges2023, author = {Glocker, Kevin and Herygers, Aaricia and Georges, Munir}, title = {Allophant: Cross-lingual Phoneme Recognition with Articulatory Attributes}, publisher = {arXiv}, address = {Ithaca}, doi = {https://doi.org/10.48550/arXiv.2306.04306}, year = {2023}, abstract = {This paper proposes Allophant, a multilingual phoneme recognizer. It requires only a phoneme inventory for crosslingual transfer to a target language, allowing for low-resource recognition. The architecture combines a compositional phone embedding approach with individually supervised phonetic attribute classifiers in a multi-task architecture. We also introduce Allophoible, an extension of the PHOIBLE database. When combined with a distance based mapping approach for grapheme-to-phoneme outputs, it allows us to train on PHOIBLE inventories directly. By training and evaluating on 34 languages, we found that the addition of multi-task learning improves the model's capability of being applied to unseen phonemes and phoneme inventories. On supervised languages we achieve phoneme error rate improvements of 11 percentage points (pp.) compared to a baseline without multi-task learning. Evaluation of zero-shot transfer on 84 languages yielded a decrease in PER of 2.63 pp. over the baseline.}, language = {en} } @unpublished{SongLyuJiangetal.2023, author = {Song, Rui and Lyu, Lingjuan and Jiang, Wei and Festag, Andreas and Knoll, Alois}, title = {V2X-Boosted Federated Learning for Cooperative Intelligent Transportation Systems with Contextual Client Selection}, publisher = {arXiv}, address = {Ithaca}, doi = {https://doi.org/10.48550/arXiv.2305.11654}, year = {2023}, abstract = {Machine learning (ML) has revolutionized transportation systems, enabling autonomous driving and smart traffic services. Federated learning (FL) overcomes privacy constraints by training ML models in distributed systems, exchanging model parameters instead of raw data. However, the dynamic states of connected vehicles affect the network connection quality and influence the FL performance. To tackle this challenge, we propose a contextual client selection pipeline that uses Vehicle-to-Everything (V2X) messages to select clients based on the predicted communication latency. The pipeline includes: (i) fusing V2X messages, (ii) predicting future traffic topology, (iii) pre-clustering clients based on local data distribution similarity, and (iv) selecting clients with minimal latency for future model aggregation. Experiments show that our pipeline outperforms baselines on various datasets, particularly in non-iid settings.}, language = {en} } @unpublished{BhanderiAgrawalElger2025, author = {Bhanderi, Savankumar and Agrawal, Shiva and Elger, Gordon}, title = {Deep Segmentation of 3+1D Radar Point Cloud for Real-Time Roadside Traffic User Detection}, titleParent = {Research Square}, publisher = {Research Square}, address = {Durham}, doi = {https://doi.org/10.21203/rs.3.rs-7222130/v1}, year = {2025}, abstract = {Smart cities rely on intelligent infrastructure to enhance road safety, optimize traffic flow, and enable vehicle-to-infrastructure (V2I) communication. A key component of such infrastructure is an efficient and real-time perception system that accurately detects diverse traffic participants. Among various sensing modalities, automotive radar is one of the best choices due to its robust performance in adverse weather and low-light conditions. However, due to low spatial resolution, traditional clustering-based approaches for radar object detection often struggle with vulnerable road user detection and nearby object separation. Hence, this paper proposes a deep learning-based 3+1D radar point cloud clustering methodology tailored for smart infrastructure-based perception applications. This approach first performs semantic segmentation of the radar point cloud, followed by instance segmentation to generate well-formed clusters with class labels using a deep neural network. It also detects single-point objects that conventional methods often miss. The described approach is developed and experimented using a smart infrastructure-based sensor setup and it performs segmentation of the point cloud in real-time. Experimental results demonstrate 95.35\% F1-macro score for semantic segmentation and 91.03\% mean average precision (mAP) at an intersection over union (IoU) threshold of 0.5 for instance segmentation. Further, the complete pipeline operates at 43.61 frames per second with a memory requirement of less than 0.7 MB on the edge device (Nvidia Jetson AGX Orin).}, language = {en} } @unpublished{IvanHirlingGrexaetal.2025, author = {Ivan, Zsanett Zsofia and Hirling, Dominik and Grexa, Istvan and Ammeling, Jonas and Micsik, Tamas and Dobra, Katalin and Kuthi, Levente and Farkas, Sukosd and Aubreville, Marc and Miczan, Vivien and Horvath, Peter}, title = {Subphase-Labeled Mitotic Dataset for AI-powered Cell Division Analysis}, publisher = {bioRxiv}, address = {Cold Spring Harbor}, doi = {https://doi.org/10.1101/2025.07.17.665280}, year = {2025}, abstract = {Mitosis detection represents a critical task in the field of digital pathology, as determination of the mitotic index (MI) plays an important role in the tumor grading and prognostic assessment of patients. Manual determination of MI is a labor-intensive and time-consuming task for practitioners with rather high interobserver variability, thus, automation has become a priority. There has been substantial progress towards creating robust mitosis detection algorithms in recent years, primarily driven by the Mitosis Domain Generalization (MIDOG) challenges. In parallel, there has been growing interest in the molecular characterization of mitosis with the goal of achieving a more comprehensive understanding of its underlying mechanisms in a subphase-specific manner. Here, we introduce a new mitotic figure dataset annotated with subphase information based on the MIDOG++ dataset as well as a previously unrepresented tumor domain to enhance the diversity and applicability of the dataset. We envision a new perspective for domain generalization by improving the performance of models with subtyping mitotic cells into the 5 main stages of normal mitosis, complemented with an atypical mitotic class. We believe that our work broadens the horizon in digital pathology: subtyping information could provide useful help for mitosis detection, while also providing promising new directions in answering biological questions, such as molecular analysis of the subphases on a single cell level.}, language = {en} } @unpublished{MohdSchmidElger2025, author = {Mohd, Zubair Akhtar and Schmid, Maximilian and Elger, Gordon}, title = {AI-Driven Point Cloud Framework for Predicting Solder Joint Reliability using 3D FEA Data}, titleParent = {Research Square}, publisher = {Research Square}, address = {Durham}, doi = {https://doi.org/10.21203/rs.3.rs-6173485/v1}, year = {2025}, abstract = {Crack propagation in solder joints remains a critical challenge affecting the thermo-mechanical reliability of electronic devices, emphasizing the need for optimized package and solder pad designs. Traditional Finite Element Analysis (FEA) methods for predicting solder joint lifespan rely heavily on manual post-processing, where high-risk regions for plastic strain accumulation are identified. However, these approaches often overlook intricate failure mechanisms, as they primarily average creep strain and correlate it with experimental lifetime data using the Coffin-Manson equation, limiting their predictive accuracy. To overcome these limitations, this study introduces a novel AI-driven framework that automates 3D FEA post-processing for surface-mounted devices (SMDs) connected to printed circuit boards (PCBs). Unlike traditional methods, this framework leverages deep learning architectures—specifically, 3D Convolutional Neural Networks (CNNs) and PointNet—to extract complex spatial features directly from 3D FEA data, eliminating the need for manual interpretation. These learned features are then mapped to experimentally measured solder joint lifetimes through fully connected neural network layers, allowing the model to capture nonlinear failure behaviours that conventional methods fail to recognize. The research focuses on crack propagation in ceramic-based high-power LED packages used in automotive lighting systems, incorporating variations in two-pad and three-pad configurations, as well as thin and thick film metallized ceramic substrates with validated FEA models. Comparative analysis shows that PointNet significantly outperforms 3D CNNs, achieving an exceptionally high correlation with experimental data (R² = 99.99\%). This AI-driven automated feature extraction and lifetime prediction approach marks a major advancement over traditional FEA-based methods, offering superior accuracy, reliability, and scalability for predicting solder joint reliability in microelectronics.}, language = {en} } @unpublished{PandeyMohdVeettiletal.2025, author = {Pandey, Amit and Mohd, Zubair Akhtar and Veettil, Nandana Kappuva and Wunderle, Bernhard and Elger, Gordon}, title = {Quantitative Kernel Estimation from Traffic Signs using Slanted Edge Spatial Frequency Response as a Sharpness Metric}, titleParent = {Research Square}, publisher = {Research Square}, address = {Durham}, doi = {https://doi.org/10.21203/rs.3.rs-6725582/v1}, year = {2025}, abstract = {The sharpness is a critical optical property of automotive cameras, measured by the Spatial Frequency Response (SFR) within the end of line (EOL) test after manufacturing. This work presents a method to estimate the blurring kernel of automotive camera for state monitoring. To achieve this, Principal Component Analysis (PCA) is performed, using synthetic kernels generated by Zemax. The PCA model is built with approximately 1300 base kernels representing spatially variant point spread functions (PSFs). This model generates kernel samples during the estimation process. Synthetic images are created by convolving the synthetic kernels with reference traffic sign images and compared with real-life data captured by an automotive camera. These synthetic data are utilized for algorithm development, and later on validation is performed on real-life data. The algorithm extracts two 45 x 45 pixels regions of interest (ROIs) containing slanted edges from the blurred image and crops matching ROIs from a reference sharp image. Each candidate kernel blurs the reference ROIs, and the resulting Spatial Frequency Response (SFR) is compared with the blurred ROIs' SFR. Differential evolution optimization minimizes the SFR difference, selecting the kernel that best matches the observed blur. The final kernel is evaluated against the true kernel for accuracy. Structural similarity index measure (SSIM) between the original and estimated blurred ROIs ranges from 0.808 to 0.945. For true vs. estimated kernels, SSIM varies from 0.92 to 0.98. Pearson correlation coefficients range from 0.84 to 0.99, Cosine similarity from 0.86 to 0.98, and mean squared error (MSE) from 1.1 x 10-5 to 8.3 x 10-5. Validation on real-life camera images shows that the SSIM between estimated ROI is 0.82 indicating a sufficient level of accuracy in kernel estimation to detect potential degradation of the camera.}, language = {en} } @unpublished{AubrevilleStathonikosDonovanetal.2023, author = {Aubreville, Marc and Stathonikos, Nikolas and Donovan, Taryn and Klopfleisch, Robert and Ganz, Jonathan and Ammeling, Jonas and Wilm, Frauke and Veta, Mitko and Jabari, Samir and Eckstein, Markus and Annuscheit, Jonas and Krumnow, Christian and Bozaba, Engin and Cayir, Sercan and Gu, Hongyan and Chen, Xiang and Jahanifar, Mostafa and Shephard, Adam and Kondo, Satoshi and Kasai, Satoshi and Kotte, Sujatha and Saipradeep, Vangala and Lafarge, Maxime W. and Koelzer, Viktor H. and Wang, Ziyue and Zhang, Yongbing and Yang, Sen and Wang, Xiyue and Breininger, Katharina and Bertram, Christof}, title = {Domain generalization across tumor types, laboratories, and species - Insights from the 2022 edition of the Mitosis Domain Generalization Challenge}, publisher = {arXiv}, address = {Ithaca}, doi = {https://doi.org/10.48550/arXiv.2309.15589}, year = {2023}, abstract = {Recognition of mitotic figures in histologic tumor specimens is highly relevant to patient outcome assessment. This task is challenging for algorithms and human experts alike, with deterioration of algorithmic performance under shifts in image representations. Considerable covariate shifts occur when assessment is performed on different tumor types, images are acquired using different digitization devices, or specimens are produced in different laboratories. This observation motivated the inception of the 2022 challenge on MItosis Domain Generalization (MIDOG 2022). The challenge provided annotated histologic tumor images from six different domains and evaluated the algorithmic approaches for mitotic figure detection provided by nine challenge participants on ten independent domains. Ground truth for mitotic figure detection was established in two ways: a three-expert consensus and an independent, immunohistochemistry-assisted set of labels. This work represents an overview of the challenge tasks, the algorithmic strategies employed by the participants, and potential factors contributing to their success. With an F1 score of 0.764 for the top-performing team, we summarize that domain generalization across various tumor domains is possible with today's deep learning-based recognition pipelines. When assessed against the immunohistochemistry-assisted reference standard, all methods resulted in reduced recall scores, but with only minor changes in the order of participants in the ranking.}, language = {en} } @unpublished{HaghoferParlakBarteletal.2023, author = {Haghofer, Andreas and Parlak, Eda and Bartel, Alexander and Donovan, Taryn and Assenmacher, Charles-Antoine and Bolfa, Pompei and Dark, Michael and Fuchs-Baumgartinger, Andrea and Klang, Andrea and J{\"a}ger, Kathrin and Klopfleisch, Robert and Merz, Sophie and Richter, Barbara and Schulman, F. Yvonne and Ganz, Jonathan and Scharinger, Josef and Aubreville, Marc and Winkler, Stephan M. and Kiupel, Matti and Bertram, Christof}, title = {Nuclear Morphometry using a Deep Learning-based Algorithm has Prognostic Relevance for Canine Cutaneous Mast Cell Tumors}, publisher = {arXiv}, address = {Ithaca}, doi = {https://doi.org/10.48550/arXiv.2309.15031}, year = {2023}, abstract = {Variation in nuclear size and shape is an important criterion of malignancy for many tumor types; however, categorical estimates by pathologists have poor reproducibility. Measurements of nuclear characteristics (morphometry) can improve reproducibility, but manual methods are time consuming. In this study, we evaluated fully automated morphometry using a deep learning-based algorithm in 96 canine cutaneous mast cell tumors with information on patient survival. Algorithmic morphometry was compared with karyomegaly estimates by 11 pathologists, manual nuclear morphometry of 12 cells by 9 pathologists, and the mitotic count as a benchmark. The prognostic value of automated morphometry was high with an area under the ROC curve regarding the tumor-specific survival of 0.943 (95\% CI: 0.889 - 0.996) for the standard deviation (SD) of nuclear area, which was higher than manual morphometry of all pathologists combined (0.868, 95\% CI: 0.737 - 0.991) and the mitotic count (0.885, 95\% CI: 0.765 - 1.00). At the proposed thresholds, the hazard ratio for algorithmic morphometry (SD of nuclear area ≥9.0μm2) was 18.3 (95\% CI: 5.0 - 67.1), for manual morphometry (SD of nuclear area ≥10.9μm2) 9.0 (95\% CI: 6.0 - 13.4), for karyomegaly estimates 7.6 (95\% CI: 5.7 - 10.1), and for the mitotic count 30.5 (95\% CI: 7.8 - 118.0). Inter-rater reproducibility for karyomegaly estimates was fair (κ = 0.226) with highly variable sensitivity/specificity values for the individual pathologists. Reproducibility for manual morphometry (SD of nuclear area) was good (ICC = 0.654). This study supports the use of algorithmic morphometry as a prognostic test to overcome the limitations of estimates and manual measurements.}, language = {en} } @unpublished{GanzMarzahlAmmelingetal.2024, author = {Ganz, Jonathan and Marzahl, Christian and Ammeling, Jonas and Richter, Barbara and Puget, Chlo{\´e} and Denk, Daniela and Demeter, Elena A. and Tabaran, Flaviu A. and Wasinger, Gabriel and Lipnik, Karoline and Tecilla, Marco and Valentine, Matthew J. and Dark, Michael and Abele, Niklas and Bolfa, Pompei and Erber, Ramona and Klopfleisch, Robert and Merz, Sophie and Donovan, Taryn and Jabari, Samir and Bertram, Christof and Breininger, Katharina and Aubreville, Marc}, title = {On the Value of PHH3 for Mitotic Figure Detection on H\&E-stained Images}, publisher = {arXiv}, address = {Ithaca}, doi = {https://doi.org/10.48550/arXiv.2406.19899}, year = {2024}, abstract = {The count of mitotic figures (MFs) observed in hematoxylin and eosin (H\&E)-stained slides is an important prognostic marker as it is a measure for tumor cell proliferation. However, the identification of MFs has a known low inter-rater agreement. Deep learning algorithms can standardize this task, but they require large amounts of annotated data for training and validation. Furthermore, label noise introduced during the annotation process may impede the algorithm's performance. Unlike H\&E, the mitosis-specific antibody phospho-histone H3 (PHH3) specifically highlights MFs. Counting MFs on slides stained against PHH3 leads to higher agreement among raters and has therefore recently been used as a ground truth for the annotation of MFs in H\&E. However, as PHH3 facilitates the recognition of cells indistinguishable from H\&E stain alone, the use of this ground truth could potentially introduce noise into the H\&E-related dataset, impacting model performance. This study analyzes the impact of PHH3-assisted MF annotation on inter-rater reliability and object level agreement through an extensive multi-rater experiment. We found that the annotators' object-level agreement increased when using PHH3-assisted labeling. Subsequently, MF detectors were evaluated on the resulting datasets to investigate the influence of PHH3-assisted labeling on the models' performance. Additionally, a novel dual-stain MF detector was developed to investigate the interpretation-shift of PHH3-assisted labels used in H\&E, which clearly outperformed single-stain detectors. However, the PHH3-assisted labels did not have a positive effect on solely H\&E-based models. The high performance of our dual-input detector reveals an information mismatch between the H\&E and PHH3-stained images as the cause of this effect.}, language = {en} } @unpublished{GanzAmmelingRosbachetal.2024, author = {Ganz, Jonathan and Ammeling, Jonas and Rosbach, Emely and Lausser, Ludwig and Bertram, Christof and Breininger, Katharina and Aubreville, Marc}, title = {Is Self-Supervision Enough? Benchmarking Foundation Models Against End-to-End Training for Mitotic Figure Classification}, publisher = {arXiv}, address = {Ithaca}, doi = {https://doi.org/10.48550/arXiv.2412.06365}, year = {2024}, abstract = {Foundation models (FMs), i.e., models trained on a vast amount of typically unlabeled data, have become popular and available recently for the domain of histopathology. The key idea is to extract semantically rich vectors from any input patch, allowing for the use of simple subsequent classification networks potentially reducing the required amounts of labeled data, and increasing domain robustness. In this work, we investigate to which degree this also holds for mitotic figure classification. Utilizing two popular public mitotic figure datasets, we compared linear probing of five publicly available FMs against models trained on ImageNet and a simple ResNet50 end-to-end-trained baseline. We found that the end-to-end-trained baseline outperformed all FM-based classifiers, regardless of the amount of data provided. Additionally, we did not observe the FM-based classifiers to be more robust against domain shifts, rendering both of the above assumptions incorrect.}, language = {en} } @unpublished{PugetGanzOstermaieretal.2024, author = {Puget, Chlo{\´e} and Ganz, Jonathan and Ostermaier, Julian and Konrad, Thomas and Parlak, Eda and Bertram, Christof and Kiupel, Matti and Breininger, Katharina and Aubreville, Marc and Klopfleisch, Robert}, title = {Deep Learning model predicts the c-Kit-11 mutational status of canine cutaneous mast cell tumors by HE stained histological slides}, publisher = {arXiv}, address = {Ithaca}, doi = {https://doi.org/10.48550/arXiv.2401.06169}, year = {2024}, abstract = {Numerous prognostic factors are currently assessed histopathologically in biopsies of canine mast cell tumors to evaluate clinical behavior. In addition, PCR analysis of the c-Kit exon 11 mutational status is often performed to evaluate the potential success of a tyrosine kinase inhibitor therapy. This project aimed at training deep learning models (DLMs) to identify the c-Kit-11 mutational status of MCTs solely based on morphology without additional molecular analysis. HE slides of 195 mutated and 173 non-mutated tumors were stained consecutively in two different laboratories and scanned with three different slide scanners. This resulted in six different datasets (stain-scanner variations) of whole slide images. DLMs were trained with single and mixed datasets and their performances was assessed under scanner and staining domain shifts. The DLMs correctly classified HE slides according to their c-Kit 11 mutation status in, on average, 87\% of cases for the best-suited stain-scanner variant. A relevant performance drop could be observed when the stain-scanner combination of the training and test dataset differed. Multi-variant datasets improved the average accuracy but did not reach the maximum accuracy of algorithms trained and tested on the same stain-scanner variant. In summary, DLM-assisted morphological examination of MCTs can predict c-Kit-exon 11 mutational status of MCTs with high accuracy. However, the recognition performance is impeded by a change of scanner or staining protocol. Larger data sets with higher numbers of scans originating from different laboratories and scanners may lead to more robust DLMs to identify c-Kit mutations in HE slides.}, language = {en} } @unpublished{AubrevilleGanzAmmelingetal.2024, author = {Aubreville, Marc and Ganz, Jonathan and Ammeling, Jonas and Kaltenecker, Christopher C. and Bertram, Christof}, title = {Model-based Cleaning of the QUILT-1M Pathology Dataset for Text-Conditional Image Synthesis}, publisher = {arXiv}, address = {Ithaca}, doi = {https://doi.org/10.48550/arXiv.2404.07676}, year = {2024}, abstract = {The QUILT-1M dataset is the first openly available dataset containing images harvested from various online sources. While it provides a huge data variety, the image quality and composition is highly heterogeneous, impacting its utility for text-conditional image synthesis. We propose an automatic pipeline that provides predictions of the most common impurities within the images, e.g., visibility of narrators, desktop environment and pathology software, or text within the image. Additionally, we propose to use semantic alignment filtering of the image-text pairs. Our findings demonstrate that by rigorously filtering the dataset, there is a substantial enhancement of image fidelity in text-to-image tasks.}, language = {en} } @unpublished{GanzMarzahlAmmelingetal.2024, author = {Ganz, Jonathan and Marzahl, Christian and Ammeling, Jonas and Rosbach, Emely and Richter, Barbara and Puget, Chlo{\´e} and Denk, Daniela and Demeter, Elena A. and Tabaran, Flaviu A. and Wasinger, Gabriel and Lipnik, Karoline and Tecilla, Marco and Valentine, Matthew J. and Dark, Michael and Abele, Niklas and Bolfa, Pompei and Erber, Ramona and Klopfleisch, Robert and Merz, Sophie and Donovan, Taryn and Jabari, Samir and Bertram, Christof and Breininger, Katharina and Aubreville, Marc}, title = {Information Mismatch in PHH3-Assisted Mitosis Annotation Leads to Interpretation Shifts in H\&E Slide Analysis}, titleParent = {Research Square}, publisher = {Research Square}, address = {Durham}, doi = {https://doi.org/10.21203/rs.3.rs-4900505/v1}, year = {2024}, abstract = {The count of mitotic figures (MFs) observed in hematoxylin and eosin (H\&E)-stained slides is an important prognostic marker, as it is a measure for tumor cell proliferation. However, the identification of MFs has a known low inter-rater agreement. In a computer-aided setting, deep learning algorithms can help to mitigate this, but they require large amounts of annotated data for training and validation. Furthermore, label noise introduced during the annotation process may impede the algorithms' performance. Unlike H\&E, where identification of MFs is based mainly on morphological features, the mitosis-specific antibody phospho-histone H3 (PHH3) specifically highlights MFs. Counting MFs on slides stained against PHH3 leads to higher agreement among raters and has therefore recently been used as a ground truth for the annotation of MFs in H\&E. However, as PHH3 facilitates the recognition of cells indistinguishable from H\&E staining alone, the use of this ground truth could potentially introduce an interpretation shift and even label noise into the H\&E-related dataset, impacting model performance. This study analyzes the impact of PHH3-assisted MF annotation on inter-rater reliability and object level agreement through an extensive multi-rater experiment. Subsequently, MF detectors, including a novel dual-stain detector, were evaluated on the resulting datasets to investigate the influence of PHH3-assisted labeling on the models' performance. We found that the annotators' object-level agreement significantly increased when using PHH3-assisted labeling (F1: 0.53 to 0.74). However, this enhancement in label consistency did not translate to improved performance for H\&E-based detectors, neither during the training phase nor the evaluation phase. Conversely, the dual-stain detector was able to benefit from the higher consistency. This reveals an information mismatch between the H\&E and PHH3-stained images as the cause of this effect, which renders PHH3-assisted annotations not well-aligned for use with H\&E-based detectors. Based on our findings, we propose an improved PHH3-assisted labeling procedure.}, language = {en} } @unpublished{AmmelingHeckerGanzetal.2023, author = {Ammeling, Jonas and Hecker, Moritz and Ganz, Jonathan and Donovan, Taryn and Klopfleisch, Robert and Bertram, Christof and Breininger, Katharina and Aubreville, Marc}, title = {Automated Volume Corrected Mitotic Index Calculation Through Annotation-Free Deep Learning using Immunohistochemistry as Reference Standard}, publisher = {arXiv}, address = {Ithaca}, doi = {https://doi.org/10.48550/arXiv.2311.08949}, year = {2023}, abstract = {The volume-corrected mitotic index (M/V-Index) was shown to provide prognostic value in invasive breast carcinomas. However, despite its prognostic significance, it is not established as the standard method for assessing aggressive biological behaviour, due to the high additional workload associated with determining the epithelial proportion. In this work, we show that using a deep learning pipeline solely trained with an annotation-free, immunohistochemistry-based approach, provides accurate estimations of epithelial segmentation in canine breast carcinomas. We compare our automatic framework with the manually annotated M/V-Index in a study with three board-certified pathologists. Our results indicate that the deep learning-based pipeline shows expert-level performance, while providing time efficiency and reproducibility.}, language = {en} } @unpublished{AmmelingSchmidtGanzetal.2022, author = {Ammeling, Jonas and Schmidt, Lars-Henning and Ganz, Jonathan and Niedermair, Tanja and Brochhausen-Delius, Christoph and Schulz, Christian and Breininger, Katharina and Aubreville, Marc}, title = {Attention-based Multiple Instance Learning for Survival Prediction on Lung Cancer Tissue Microarrays}, publisher = {arXiv}, address = {Ithaca}, doi = {https://doi.org/10.48550/arXiv.2212.07724}, year = {2022}, abstract = {Attention-based multiple instance learning (AMIL) algorithms have proven to be successful in utilizing gigapixel whole-slide images (WSIs) for a variety of different computational pathology tasks such as outcome prediction and cancer subtyping problems. We extended an AMIL approach to the task of survival prediction by utilizing the classical Cox partial likelihood as a loss function, converting the AMIL model into a nonlinear proportional hazards model. We applied the model to tissue microarray (TMA) slides of 330 lung cancer patients. The results show that AMIL approaches can handle very small amounts of tissue from a TMA and reach similar C-index performance compared to established survival prediction methods trained with highly discriminative clinical factors such as age, cancer grade, and cancer stage.}, language = {en} } @unpublished{StoeveAubrevilleOetteretal.2018, author = {Stoeve, Maike and Aubreville, Marc and Oetter, Nicolai and Knipfer, Christian and Neumann, Helmut and Stelzle, Florian and Maier, Andreas}, title = {Motion Artifact Detection in Confocal Laser Endomicroscopy Images}, publisher = {arXiv}, address = {Ithaca}, doi = {https://doi.org/10.48550/arXiv.1711.01117}, year = {2018}, abstract = {Confocal Laser Endomicroscopy (CLE), an optical imaging technique allowing non-invasive examination of the mucosa on a (sub)- cellular level, has proven to be a valuable diagnostic tool in gastroenterology and shows promising results in various anatomical regions including the oral cavity. Recently, the feasibility of automatic carcinoma detection for CLE images of sufficient quality was shown. However, in real world data sets a high amount of CLE images is corrupted by artifacts. Amongst the most prevalent artifact types are motion-induced image deteriorations. In the scope of this work, algorithmic approaches for the automatic detection of motion artifact-tainted image regions were developed. Hence, this work provides an important step towards clinical applicability of automatic carcinoma detection. Both, conventional machine learning and novel, deep learning-based approaches were assessed. The deep learning-based approach outperforms the conventional approaches, attaining an AUC of 0.90.}, language = {en} } @unpublished{AubrevillePanSievertetal.2023, author = {Aubreville, Marc and Pan, Zhaoya and Sievert, Matti and Ammeling, Jonas and Ganz, Jonathan and Oetter, Nicolai and Stelzle, Florian and Frenken, Ann-Kathrin and Breininger, Katharina and Goncalves, Miguel}, title = {Few Shot Learning for the Classification of Confocal Laser Endomicroscopy Images of Head and Neck Tumors}, publisher = {arXiv}, address = {Ithaca}, doi = {https://doi.org/10.48550/arXiv.2311.07216}, year = {2023}, abstract = {The surgical removal of head and neck tumors requires safe margins, which are usually confirmed intraoperatively by means of frozen sections. This method is, in itself, an oversampling procedure, which has a relatively low sensitivity compared to the definitive tissue analysis on paraffin-embedded sections. Confocal laser endomicroscopy (CLE) is an in-vivo imaging technique that has shown its potential in the live optical biopsy of tissue. An automated analysis of this notoriously difficult to interpret modality would help surgeons. However, the images of CLE show a wide variability of patterns, caused both by individual factors but also, and most strongly, by the anatomical structures of the imaged tissue, making it a challenging pattern recognition task. In this work, we evaluate four popular few shot learning (FSL) methods towards their capability of generalizing to unseen anatomical domains in CLE images. We evaluate this on images of sinunasal tumors (SNT) from five patients and on images of the vocal folds (VF) from 11 patients using a cross-validation scheme. The best respective approach reached a median accuracy of 79.6\% on the rather homogeneous VF dataset, but only of 61.6\% for the highly diverse SNT dataset. Our results indicate that FSL on CLE images is viable, but strongly affected by the number of patients, as well as the diversity of anatomical patterns.}, language = {en} } @unpublished{PerniasRampasAubreville2023, author = {Pernias, Pablo and Rampas, Dominic and Aubreville, Marc}, title = {W{\"u}rstchen: Efficient Pretraining of Text-to-Image Models}, publisher = {arXiv}, address = {Ithaca}, doi = {https://doi.org/10.48550/arXiv.2306.00637}, year = {2023}, abstract = {We introduce Wuerstchen, a novel technique for text-to-image synthesis that unites competitive performance with unprecedented cost-effectiveness and ease of training on constrained hardware. Building on recent advancements in machine learning, our approach, which utilizes latent diffusion strategies at strong latent image compression rates, significantly reduces the computational burden, typically associated with state-of-the-art models, while preserving, if not enhancing, the quality of generated images. Wuerstchen achieves notable speed improvements at inference time, thereby rendering real-time applications more viable. One of the key advantages of our method lies in its modest training requirements of only 9,200 GPU hours, slashing the usual costs significantly without compromising the end performance. In a comparison against the state-of-the-art, we found the approach to yield strong competitiveness. This paper opens the door to a new line of research that prioritizes both performance and computational accessibility, hence democratizing the use of sophisticated AI technologies. Through Wuerstchen, we demonstrate a compelling stride forward in the realm of text-to-image synthesis, offering an innovative path to explore in future research.}, language = {en} } @unpublished{MertesKarleHuberetal.2022, author = {Mertes, Silvan and Karle, Christina and Huber, Tobias and Weitz, Katharina and Schlagowski, Ruben and Andr{\´e}, Elisabeth}, title = {Alterfactual Explanations - The Relevance of Irrelevance for Explaining AI Systems}, publisher = {arXiv}, address = {Ithaca}, doi = {https://doi.org/10.48550/arXiv.2207.09374}, year = {2022}, abstract = {Explanation mechanisms from the field of Counterfactual Thinking are a widely-used paradigm for Explainable Artificial Intelligence (XAI), as they follow a natural way of reasoning that humans are familiar with. However, all common approaches from this field are based on communicating information about features or characteristics that are especially important for an AI's decision. We argue that in order to fully understand a decision, not only knowledge about relevant features is needed, but that the awareness of irrelevant information also highly contributes to the creation of a user's mental model of an AI system. Therefore, we introduce a new way of explaining AI systems. Our approach, which we call Alterfactual Explanations, is based on showing an alternative reality where irrelevant features of an AI's input are altered. By doing so, the user directly sees which characteristics of the input data can change arbitrarily without influencing the AI's decision. We evaluate our approach in an extensive user study, revealing that it is able to significantly contribute to the participants' understanding of an AI. We show that alterfactual explanations are suited to convey an understanding of different aspects of the AI's reasoning than established counterfactual explanation methods.}, language = {en} } @unpublished{HaryantoVaculin2025, author = {Haryanto, Aditya and Vaculin, Ondrej}, title = {YoFlow Method for Scenario Based Automatic Accident Detection}, publisher = {TechRxiv}, address = {Piscataway}, doi = {https://doi.org/10.36227/techrxiv.175099959.99453472/v1}, year = {2025}, abstract = {Recent advances in sensor and computing technologies have enabled roadside units (RSUs) to not only monitor traffic flow but also process data in real time to improve road safety. However, leveraging RSUs for proactive accident detection remains a challenging and underexplored task, partly due to the lack of diverse accident data. To address this, this study proposes two key contributions: (i) a scenario based synthetic data generation framework, and (ii) YoFlow, a novel system for vehicle-to-vehicle accident detection from a simulated RSU camera perspective. The proposed framework leverages the PEGASUS methodology and the BeamNG.tech simulation platform to create the SB-SIF dataset, which includes five representative intersection crash scenarios derived from German accident data. SB-SIF dataset contains 914 crash videos, 123 near-miss events, and 924 normal traffic instances and is publicly available at: https://doi.org/10.5281/zenodo.15267252. The proposed YoFlow system identifies accidents by analyzing temporal variations in vehicle speed vectors, using YOLO for vehicle classification and CUDA-accelerated dense optical flow to capture abrupt motion changes. Extracted features are processed and classified using an XGBoost model, achieving 94\% recall and 90\% precision in accident detection.}, language = {en} } @unpublished{SuemerHuberDuongetal.2025, author = {S{\"u}mer, {\"O}mer and Huber, Tobias and Duong, Dat and Ledgister Hanchard, Suzanna E. and Conati, Cristina and Andr{\´e}, Elisabeth and Solomon, Benjamin D. and Waikel, Rebekah L.}, title = {Evaluation of a Deep Learning and XAI based Facial Phenotyping Tool for Genetic Syndromes: A Clinical User Study}, publisher = {medRxiv}, address = {Cold Spring Harbor}, doi = {https://doi.org/10.1101/2025.06.08.25328588}, year = {2025}, abstract = {Artificial intelligence (AI) tools are increasingly employed in clinical genetics to assist in diagnosing genetic conditions by assessing photographs of patients. For medical uses of AI, explainable AI (XAI) methods offer a promising approach by providing interpretable outputs, such as saliency maps and region relevance visualizations. XAI has been discussed as important for regulatory purposes and to enable clinicians to better understand how AI tools work in practice. However, the real-world effects of XAI on clinician performance, confidence, and trust remain underexplored. This study involved a web-based user experiment with 31 medical geneticists to assess the impact of AI-only diagnostic assistance compared to XAI-supported diagnostics. Participants were randomly assigned to either group and completed diagnostic tasks with 18 facial images of individuals with known genetic syndromes and unaffected individuals, before and after experiencing the AI outputs. The results show that both AI-only and XAI approaches improved diagnostic accuracy and clinician confidence. The effects varied according to the accuracy of AI predictions and the clarity of syndromic features (sample difficulty). While AI support was viewed positively, users approached XAI with skepticism. Interestingly, we found a positive correlation between diagnostic improvement and XAI intervention. Although XAI support did not significantly enhance overall performance relative to AI alone, it prompted users to critically evaluate images with false predictions and influenced their confidence levels. These findings highlight the complexities of trust, perceived usefulness, and interpretability in AI-assisted diagnostics, with important implications for developing and implementing clinical decision-support tools in facial phenotyping for rare genetic diseases.}, language = {en} } @unpublished{BattahiChbaniNiederlaenderetal.2025, author = {Battahi, Fouad and Chbani, Zaki and Niederl{\"a}nder, Simon and Riahi, Hassan}, title = {Asymptotic behavior of the Arrow-Hurwicz differential system with Tikhonov regularization}, publisher = {arXiv}, address = {Ithaca}, doi = {https://doi.org/10.48550/arXiv.2411.17656}, year = {2025}, abstract = {In a real Hilbert space setting, we investigate the asymptotic behavior of the solutions of the classical Arrow-Hurwicz differential system combined with Tikhonov regularizing terms. Under some newly proposed conditions on the Tikhonov terms involved, we show that the solutions of the regularized Arrow-Hurwicz differential system strongly converge toward the element of least norm within its set of zeros. Moreover, we provide fast asymptotic decay rate estimates for the so-called primal-dual gap function and the norm of the solutions' velocity. If, in addition, the Tikhonov regularizing terms are decreasing, we provide some refined estimates in the sense of an exponentially weighted moving average. Under the additional assumption that the governing operator of the Arrow-Hurwicz differential system satisfies a reverse Lipschitz condition, we further provide a fast rate of strong convergence of the solutions toward the unique zero. We conclude our study by deriving the corresponding decay rate estimates with respect to the so-called viscosity curve. Numerical experiments illustrate our theoretical findings.}, language = {en} } @unpublished{KoenigPaulaQuaschneretal.2023, author = {K{\"o}nig, Thomas and Paula, Daniel and Quaschner, Stefan and Schweiger, Hans-Georg}, title = {Influence of Blind Spot Assistance Systems in Heavy Commercial Vehicles on Accident Reconstruction}, publisher = {Preprints}, address = {Basel}, doi = {https://doi.org/10.20944/preprints202311.1308.v1}, year = {2023}, abstract = {Accidents between right-turning commercial vehicles and crossing vulnerable road users (VRU) in urban environments often lead to serious or fatal injuries and therefore play a significant role in forensic accident analysis. To reduce the risk of accidents, blind spot assistance systems have been installed in commercial vehicles for several years, among other things, to detect VRUs and warn the driver in time. However, since such systems cannot reliably prevent all turning accidents, an investigation by experts must clarify how the accident occurred and to what extent the blind spot assistance system influenced the course of the accident. The occurrence of the acoustic warning message can be defined as an objective reaction prompt for the driver, so that the blind spot assistance system can significantly influence the avoidability assessment. In order to be able to integrate the system into forensic accident analysis, a precise knowledge of how the system works and its limitations is required. For this purpose, tests with different systems and accident constellations were conducted and evaluated. It was found that the type of sensor used for the assistance systems has a great influence on the system's performance. The lateral distance between the right side of the commercial vehicle and the VRU as well as obstacles between them and the speed difference can take great influence on the reliability of the assistance system. Depending on the concrete time of the system's warning signal the accident can be avoided or not by the driver when reacting on this signal.}, language = {en} } @unpublished{KettelgerdesHillmannHirmeretal.2023, author = {Kettelgerdes, Marcel and Hillmann, Tjorven and Hirmer, Thomas and Erdogan, H{\"u}seyin and Wunderle, Bernhard and Elger, Gordon}, title = {Accelerated Real-Life (ARL) Testing and Characterization of Automotive LiDAR Sensors to facilitate the Development and Validation of Enhanced Sensor Models}, publisher = {arXiv}, address = {Ithaca}, doi = {https://doi.org/10.48550/arXiv.2312.04229}, year = {2023}, abstract = {In the realm of automated driving simulation and sensor modeling, the need for highly accurate sensor models is paramount for ensuring the reliability and safety of advanced driving assistance systems (ADAS). Hence, numerous works focus on the development of high-fidelity models of ADAS sensors, such as camera, Radar as well as modern LiDAR systems to simulate the sensor behavior in different driving scenarios, even under varying environmental conditions, considering for example adverse weather effects. However, aging effects of sensors, leading to suboptimal system performance, are mostly overlooked by current simulation techniques. This paper introduces a cutting-edge Hardware-in-the-Loop (HiL) test bench designed for the automated, accelerated aging and characterization of Automotive LiDAR sensors. The primary objective of this research is to address the aging effects of LiDAR sensors over the product life cycle, specifically focusing on aspects such as laser beam profile deterioration, output power reduction and intrinsic parameter drift, which are mostly neglected in current sensor models. By that, this proceeding research is intended to path the way, not only towards identifying and modeling respective degradation effects, but also to suggest quantitative model validation metrics.}, language = {en} } @unpublished{PagonisHartungWuetal.2024, author = {Pagonis, Panagiotis and Hartung, Kai and Wu, Di and Georges, Munir and Gr{\"o}ttrup, S{\"o}ren}, title = {Analysis of Knowledge Tracing performance on synthesised student data}, publisher = {arXiv}, address = {Ithaca}, doi = {https://doi.org/10.48550/arXiv.2401.16832}, year = {2024}, abstract = {Knowledge Tracing (KT) aims to predict the future performance of students by tracking the development of their knowledge states. Despite all the recent progress made in this field, the application of KT models in education systems is still restricted from the data perspectives: 1) limited access to real life data due to data protection concerns, 2) lack of diversity in public datasets, 3) noises in benchmark datasets such as duplicate records. To resolve these problems, we simulated student data with three statistical strategies based on public datasets and tested their performance on two KT baselines. While we observe only minor performance improvement with additional synthetic data, our work shows that using only synthetic data for training can lead to similar performance as real data.}, language = {en} } @unpublished{SuciuBock2025, author = {Suciu, Tobias and Bock, J{\"u}rgen}, title = {A Practical Approach to Multivariate Time Series Anomaly Detection in Automotive Bus Systems Testing}, titleParent = {Research Square}, publisher = {Research Square}, address = {Durham}, doi = {https://doi.org/10.21203/rs.3.rs-7778102/v1}, year = {2025}, abstract = {The increasing complexity of modern vehicles and their testing procedures generates vast amounts of multivariate time series data, making manual anomaly detection during automotive testing increasingly challenging. This article investigates the application of deep learning algorithms for automated anomaly detection in automotive bus data collected during dynamic driving scenarios. Three distinct architectures are implemented and compared: a CNN-based forecasting approach (DeepAnT), an LSTM-based model (LSTM-AD), and a Convolutional Autoencoder (CAE). Real-world driving data collected across various scenarios, ranging from normal operation to extreme maneuvers, is employed. Through evaluation across seven distinct test scenarios, findings reveal that while each architecture demonstrates specific strengths, their effectiveness varies significantly based on anomaly type and driving context. DeepAnT shows the most consistent performance across different scenarios, while LSTM-AD achieves superior detection capability for complex temporal patterns, particularly in scenarios involving coordinated changes across multiple features. The CAE excels at identifying pronounced deviations but shows limitations in detecting subtle anomalies. This study demonstrates that while deep learning models effectively detect anomalies in automotive time series data, their practical implementation requires careful consideration of specific use cases, emphasizing the critical role of data preprocessing and threshold calculation in ensuring reliable anomaly detection.}, language = {en} } @unpublished{OkumusRamicKugele2025, author = {Okumus, Fazli Faruk and Ramic, Amra and Kugele, Stefan}, title = {A Systematic Mapping Study on Contract-based Software Design for Dependable Systems}, publisher = {arXiv}, address = {Ithaca}, doi = {https://doi.org/10.48550/arXiv.2505.07542}, year = {2025}, abstract = {Background: Contract-based Design (CbD) is a valuable methodology for software design that allows annotation of code and architectural components with contracts, thereby enhancing clarity and reliability in software development. It establishes rules that outline the behaviour of software components and their interfaces and interactions. This modular approach enables the design process to be segmented into smaller, independently developed, tested, and verified system components, ultimately leading to more robust and dependable software. Aim: Despite the significance and well-established theoretical background of CbD, there is a need for a comprehensive systematic mapping study for reliable software systems. Our study provides an evidence-based overview of a method and demonstrates its practical feasibility. Method: To conduct this study, we systematically searched three different databases using specially formulated queries, which initially yielded 1,221 primary studies. After voting, we focused on 288 primary studies for more detailed analysis. Finally, a collaborative review allowed us to gather relevant evidence and information to address our research questions. Results: Our findings suggest potential avenues for future research trajectories in CbD, emphasising its role in improving the dependability of software systems. We highlight maturity levels across different domains and identify areas that may benefit from further research. Conclusion: Although CbD is a well-established software design approach, a more comprehensive literature review is needed to clarify its theoretical state about dependable systems. Our study addresses this gap by providing a detailed overview of CbD from various perspectives, identifying key gaps, and suggesting future research directions.}, language = {en} } @unpublished{DoenmezTejerodelaPiedraKloseetal.2026, author = {D{\"o}nmez, {\"O}mer and Tejero de la Piedra, Ricardo and Klose, Simona and Riolet, Matthieu and Rozek, Lukas and Vaculin, Ondrej and Hach, Christian}, title = {Approach for Passive Safety Assessment of Rearward-Sitting Occupants}, publisher = {Zenodo}, address = {Genf}, doi = {https://doi.org/10.5281/zenodo.18338626}, year = {2026}, abstract = {The introduction of highly automated vehicles (HAVs) will allow vehicle occupants to take advantage of new seating configurations, such as sitting rearward in the first row. One critical aspect of assessing occupant safety during high-speed impacts is the lack of a dedicated safety framework for rearward-facing passengers in the first row. This paper introduces a method to develop new assessment criteria for these novel seat configurations. Thus, this research presents some preliminary results of rearward-facing occupant injury biomechanics analyses carried out employing a variety of anthropomorphic test devices (ATDs) and the VIVA+ 50M human body model (HBM), restrained with different belt configurations and considering different seat typologies. It reviews the suitability of 50th percentile male ATDs to capture a biofidelic engagement with the seat structure and belt system and evaluates the reaction loads on the occupant, along with the energy management resulting from seat back rotational stiffness and energy-absorbing foams layered behind the seat cushion. Based on the results, the THOR-AV-50M is a suitable candidate for further biofidelity analysis. Torso occupant loads can be effectively reduced utilizing seat back rotation but pelvis load management requires further studies.}, language = {en} } @unpublished{SalamatMatternOlzemetal.2025, author = {Salamat, Babak and Mattern, Dominik and Olzem, Sebastian-Sven and Elsbacher, Gerhard and Seidel, Christian and Tonello, Andrea M.}, title = {GMP3: Learning-Driven, Bellman-Guided Trajectory Planning for UAVs in Real-Time on SE(3)}, publisher = {arXiv}, address = {Ithaca}, doi = {https://doi.org/10.48550/arXiv.2509.21264}, year = {2025}, abstract = {We propose GMP3, a multiphase global path planning framework that generates dynamically feasible three-dimensional trajectories for unmanned aerial vehicles (UAVs) operating in cluttered environments. The framework extends traditional path planning from Euclidean position spaces to the Lie group SE(3), allowing joint learning of translational motion and rotational dynamics. A modified Bellman-based operator is introduced to support reinforcement learning (RL) policy updates while leveraging prior trajectory information for improved convergence. GMP3 is designed as a distributed framework in which agents influence each other and share policy information along the trajectory: each agent refines its assigned segment and shares with its neighbors via a consensus-based scheme, enabling cooperative policy updates and convergence toward a path shaped globally even under kinematic constraints. We also propose DroneManager, a modular ground control software that interfaces the planner with real UAV platforms via the MAVLink protocol, supporting real-time deployment and feedback. Simulation studies and indoor flight experiments validate the effectiveness of the proposed method in constrained 3D environments, demonstrating reliable obstacle avoidance and smooth, feasible trajectories across both position and orientation. The open-source implementation is available at https://github.com/Domattee/DroneManager.}, language = {en} }