Although Vision Language Models (VLMs) have seen tremendous progress across all kinds of use cases, they still fall behind in answering questions regarding diagrams compared to photos. Although progress has been made in the area of bar charts, line charts and similar diagrams, there is still few research concerned with other types of diagrams, e.g. in the computer science domain. We identified a gap in research on visual question answering on UML class diagrams. Our objective is to fill the gap by analyzing the performance of popular open-weight VLMs on a self-constructed benchmark for visual question answering based on UML class diagrams which is both challenging and manageable. We further construct a large-scale training dataset with 16.000 image-question-an-swer triples based on real software repositories on GitHub. We focus on Java-based repositories and filter for project sizes that are small enough to fit on a single diagram with 4000x4000 pixels maximum in a readable manner. We ask questions based on 18 question templates. We show that a LoRA-based finetune of Qwen 2.5 VL 7B easily outperforms Qwen 3.5 27B, which is a recent and well-performing VLM in many other benchmarks.

Physical Unclonable Functions (PUFs) exploit inherent manufacturing variations in electronic devices to create unique hardware fingerprints, which are typically used as a source of randomness for generating device-specific cryptographic keys. In this work, however, we explore the possibility of repurposing PUFs as a secure storage mechanism for pre-existing keys. First, we evaluate existing RH-PUFs, introduce five novel RH-PUF constructions, and compare all of them with respect to their suitability for cryptographic key generation. Second, we propose DRAmGON, a novel, unconventional PUF-based approach that enables the direct encoding of cryptographic keys into Rowhammer-susceptible DRAM, thereby allowing persistent data storage in volatile memory. Our experiments using FlippyRAM show that 81 % of the tested systems exhibit Rowhammer susceptibility, establishing RH-PUFs as a viable hardware-based security primitive. 

Background

Methods

Results

Conclusions

Remote latency side channels reveal sensitive information by only observing latency variations in the attacker’s traffic with the victim. While these attacks are easy to mount and scale, there is no practical defense. Considering the more powerful attacker-in-the-middle scenarios, available mitigations require the cooperation of all communication partners for protection, and cause prohibitive overheads.

In this paper, we propose AckwardDelay, a new unilateral, purely client-side, and lightweight defense against remote latency side channels. In detail, we piece-wise apply constant-time principles on the latency side channel and computationally show that a fully remote attacker requires more than 250 samples to reduce the initial search space to below 1% with our recommended parameters, even in a scenario favoring the attacker. Based on our proof-of-concept AckwardDelay implementation with less than 1000 lines of code on Linux, we demonstrate that the accuracy of website- and video-fingerprinting attacks is reduced to random guessing in practice. With an increase of only 5.55% on website-loading times and a reduction of only 0.51% in transfer rates, we conclude that AckwardDelay is a practical, lightweight, and effective mitigation applicable to the vast number of client systems such as smart phones, tablets, and laptops.

This paper addresses the challenge of real-time tool condition monitoring in tapping processes using machine learning techniques, with a focus on cross-material generalization and robust fault detection. The study leverages a historical dataset from 1998, comprising 2,195 tapping experiments on two steel alloys - 16 MnCr 5 and 42 CrMo 4 - monitoring torque (Mz) signals to predict binary quality outcomes (good/bad) based on defined quality criteria. To overcome limitations in prior work, the authors introduce a feature extraction method that captures both amplitude and duration characteristics across distinct phases of the torque signal,. The evaluation framework includes increasingly challenging train/test splits: random, run-wise holdout, and cross-material (training on 16 MnCr 5, testing on 42 CrMo 4), enabling assessment of real-world generalizability.

Multiple machine learning models are tested using both raw time-series data (after cleaning and normalization) and engineered features. Results show Matthews Correlation Coefficients (MCC) of 0.40 - 0.41 under cross-material testing-indicating moderate but meaningful generalization across materials with different mechanical properties. This performance level suggests that fundamental physical regularities in successful tapping produce consistent torque signatures, enabling transferable detection of anomalies without retraining. Findings support the feasibility of plug-and-play monitoring systems in agile manufacturing environments, where minimal setup and broad applicability are essential.

Download-Link: Revisiting 1998 Torque Data: A Machine Learning Analysis of Time Series Data for Tapping Experiments

Additive Manufacturing (AM) is increasingly adopted in the aerospace industry, as benefits like resource efficiency are complemented by distributed manufacturing possibilities that enhance supply chain resilience. However, replacing con ventional, established manufacturing methods with Laser Powder Bed Fusion in a highly regulated domain such as civil aviation comes at the price of increased requirements and thus costs for certification and quality assurance, which limit the attractiveness of AM. This paper presents a new data-based certification platform using Machine Learning (ML), a subdomain of artificial intelligence (AI), to enable faster and more cost-efficient design and manufacturing approval for additively manufactured aircraft components. The platform connects all relevant stakeholders and guides them through the certification process. As data sharing across different stakeholders and ML applications are central to the platform, a data governance concept aligned with European legislation, based on project-specific closed groups comprising direct supplier-customer relationships was developed. In addition, a compatible platform business model is described, presenting the roles of stakeholders and their respective value contributions. To this end, a deep dive into the landscape of current certification approaches and requirements was conducted and the impact of AM and the general use of AI on aircraft component certification was evaluated from technical, regulatory, legal, and economic perspectives

We investigate the effectiveness of anti-phishing support systems through a quantitative study involving 453 participants. To this end, we developed a tool that allows participants to immerse themselves in a realistic setting, tasked with classifying emails as either phishing or legitimate, while being assisted by support systems. Despite the prevalence of support systems in webmailers and email clients, our results indicate no significant difference in correctly assessing emails of varying difficulty between these systems and the control group. We found a minor negative effect of the support system that uses tooltips compared to other support systems. In the subsequent survey, we found that the support systems are appreciated and considered helpful by users, as supported by the results of the UEQ-S, even if they have no observable effect. Email context, such as the contact list, as well as hovering over the links, had stronger effects on the classification than the tested support systems.