Search

Research software is an import output of research and must be published according to the FAIR Principles for Research Software. This can be achieved by publishing software with metadata under a persistent identifier. HERMES is a tool that leverages continuous integration to automate the publication of software with rich metadata. In this work, we describe the HERMES workflow itself, and how to extend it to meet the needs of specific research software metadata or infrastructure. We introduce the HERMES plugin architecture and provide the example of creating a new HERMES plugin that harvests metadata from a metadata source in source code repositories. We show how to use HERMES as an end user, both via the command line interface, and as a step in a continuous integration pipeline. Finally, we report three informal case studies whose results provide a preliminary evaluation of the feasibility and applicability of HERMES workflows, and the extensibility of the hermes software package., Comment: 17 pages, 2 figures, 2 tables, submitted to a special issue of Electronic Communications of the EASST collecting submissions of deRSE24, Conference for Research Software Engineers

While generative AI (GenAI) offers countless possibilities for creative and productive tasks, artificially generated media can be misused for fraud, manipulation, scams, misinformation campaigns, and more. To mitigate the risks associated with maliciously generated media, forensic classifiers are employed to identify AI-generated content. However, current forensic classifiers are often not evaluated in practically relevant scenarios, such as the presence of an attacker or when real-world artifacts like social media degradations affect images. In this paper, we evaluate state-of-the-art AI-generated image (AIGI) detectors under different attack scenarios. We demonstrate that forensic classifiers can be effectively attacked in realistic settings, even when the attacker does not have access to the target model and post-processing occurs after the adversarial examples are created, which is standard on social media platforms. These attacks can significantly reduce detection accuracy to the extent that the risks of relying on detectors outweigh their benefits. Finally, we propose a simple defense mechanism to make CLIP-based detectors, which are currently the best-performing detectors, robust against these attacks.

System prompts that include detailed instructions to describe the task performed by the underlying large language model (LLM) can easily transform foundation models into tools and services with minimal overhead. Because of their crucial impact on the utility, they are often considered intellectual property, similar to the code of a software product. However, extracting system prompts is easily possible by using prompt injection. As of today, there is no effective countermeasure to prevent the stealing of system prompts and all safeguarding efforts could be evaded with carefully crafted prompt injections that bypass all protection mechanisms. In this work, we propose an alternative to conventional system prompts. We introduce prompt obfuscation to prevent the extraction of the system prompt while maintaining the utility of the system itself with only little overhead. The core idea is to find a representation of the original system prompt that leads to the same functionality, while the obfuscated system prompt does not contain any information that allows conclusions to be drawn about the original system prompt. We implement an optimization-based method to find an obfuscated prompt representation while maintaining the functionality. To evaluate our approach, we investigate eight different metrics to compare the performance of a system using the original and the obfuscated system prompts, and we show that the obfuscated version is constantly on par with the original one. We further perform three different deobfuscation attacks and show that with access to the obfuscated prompt and the LLM itself, we are not able to consistently extract meaningful information. Overall, we showed that prompt obfuscation can be an effective method to protect intellectual property while maintaining the same utility as the original system prompt.

There is a broad class of networks which connect inputs to outputs. We provide a strong theoretical foundation for crossover across this class and connect it to informativeness, a measure of the connectedness of inputs to outputs. We define Input/Output Directed Graphs (or IOD Graphs) as graphs with nodes $N$ and directed edges $E$, where $N$ contains (a) a set of "input nodes" $I \subset N$, where each $i \in I$ has no incoming edges and any number of outgoing edges, and (b) a set of "output nodes" $O \subset N$, where each $o \in O$ has no outgoing edges and any number of incoming edges, and $I\cap O = \emptyset$. We define informativeness, which involves the connections via directed paths from the input nodes to the output nodes: A partially informative IOD Graph has at least one path from an input to an output, a very informative IOD Graph has a path from every input to some output, and a fully informative IOD Graph has a path from every input to every output. A perceptron is an example of an IOD Graph. If it has non-zero weights and any number of layers, it is fully informative. As links are removed (assigned zero weight), the perceptron might become very, partially, or not informative. We define a crossover operation on IOD Graphs in which we find subgraphs with matching sets of forward and backward directed links to "swap." With this operation, IOD Graphs can be subject to evolutionary computation methods. We show that fully informative parents may yield a non-informative child. We also show that under conditions of contiguousness and the no dangling nodes condition, crossover compatible, partially informative parents yield partially informative children, and very informative input parents with partially informative output parents yield very informative children. However, even under these conditions, full informativeness may not be retained.

CNNs, most notably the UNet, are the default architecture for biomedical segmentation. Transformer-based approaches, such as UNETR, have been proposed to replace them, benefiting from a global field of view, but suffering from larger runtimes and higher parameter counts. The recent Vision Mamba architecture offers a compelling alternative to transformers, also providing a global field of view, but at higher efficiency. Here, we introduce ViM-UNet, a novel segmentation architecture based on it and compare it to UNet and UNETR for two challenging microscopy instance segmentation tasks. We find that it performs similarly or better than UNet, depending on the task, and outperforms UNETR while being more efficient. Our code is open source and documented at https://github.com/constantinpape/torch-em/blob/main/vimunet.md., Comment: Published in MIDL 2024

Informing drivers about the predicted state of upcoming traffic lights is considered a key solution to reduce unneeded energy expenditure and dilemma zones at intersections. However, newer traffic lights can react to traffic demand, resulting in spontaneous switching behavior and poor predictability. To assess whether future traffic light assistance services are viable, it is crucial to understand how strongly predictability is affected by such spontaneous switching behavior. Previous studies have so far only reported percentages of adaptivity-capable traffic lights, but the actual switching behavior has not been measured. Addressing this research gap, we conduct a large-scale predictability evaluation based on 424 million recorded switching cycles over four weeks for 18,009 individual traffic lights in Hamburg. Two characteristics of predictability are studied: cycle discrepancy and wait time diversity. Results indicate that fewer traffic lights exhibit hard-to-predict switching behavior than suggested by previous work, considering a reported number of 90.7% adaptive traffic lights in Hamburg. Contrasting previous work, we find that not all traffic lights capable of adaptiveness may necessarily exhibit low predictability. We critically review these results and derive avenues for future research., Comment: Accepted for publication at IEEE IV 2024

In this paper we discuss the design and implementation of a plasma display panel for a wide field of view desktop virtual reality environment. Present commercial plasma displays are not designed with virtual reality in mind, leading to several problems in generating stereo imagery and obtaining good tracking information. Although we developed solutions for a number of these problems, the limitations of the system preclude its current use in practical applications, and point to issues that must be resolved for flat panel displays to be useful for VR., Comment: 6 pages, 5 figures, 3rd International Immersive Projection Technology Workshop (Stuttgart, Germany, 10-11 May 1999)

As part of the iGrid Research Demonstration at INET 2000, we created two Virtual Cultural Heritage environments - "Virtual Harlem" and "Shared Miletus". The purpose of these applications was to explore possibilities in using the combination of high-speed international networks and virtual reality (VR) displays for cultural heritage education. Our ultimate goal is to enable the construction of tele-immersive museums and classes. In this paper we present an overview of the infrastructure used for these applications, and some details of their construction., Comment: 9 pages, 4 figures, INET 2001 (Stockholm, Sweden, 5-8 June 2001)

We present a collection of heuristics and simple tests for evaluating the quality of a projection-based virtual reality display. A typical VR system includes numerous potential sources of error. By understanding the characteristics of a correctly working system, and the types of errors that are likely to occur, users can quickly determine if their display is inaccurate and what components may need correction., Comment: 7 pages, 6 figures, 4th International Immersive Projection Technology Workshop (Ames, Iowa, 19-20 June 2000)

We describe a system for visualization and editing of data in a computational chemistry environment. The system is a collaborative tool allowing researchers using virtual reality and/or desktop computer displays to work together on results of the Shake-and-Bake structure determination application., Comment: 6 pages, 2 figures, Proceedings of 2nd INTUITION International Workshop, Paris, France, 24-25 Nov 2005

Complex interelement trends among magmatic IIIF iron meteorites are difficult to explain by fractional crystallization and have raised uncertainty about their genetic relationships. Nucleosynthetic Mo isotope anomalies provide a powerful tool to assess if individual IIIF irons are related to each other. However, while trace-element data are available for all nine IIIF irons, Mo isotopic data are limited to three samples. We present Mo isotopic data for all but one IIIF irons that help assess the genetic relationships among these irons, together with new Mo and W isotopic data for Fitzwater Pass (classified IIIF), and the Zinder pallasite (for which a cogenetic link with IIIF irons has been proposed). After correction for cosmic-ray exposure, the Mo isotopic compositions of the IIIF irons are identical within uncertainty and confirm their belonging to carbonaceous chondrite-type (CC) meteorites. The mean Mo isotopic composition of Group IIIF overlaps those Groups IIF and IID, but a common parent body for these groups is ruled out based on distinct trace element systematics. The new Mo isotopic data do not argue against a single parent body for the IIIF irons, and suggest a close genetic link among these samples. By contrast, Fitzwater Pass has distinct Mo and W isotopic compositions, identical to those of some non-magmatic IAB irons. The Mo and W isotope data for Zinder indicate that this meteorite is not related to IIIF irons, but belongs to the non-carbonaceous (NC) type and has the same Mo and W isotopic composition as main-group pallasites., Comment: Accepted for publication in Meteoritics & Planetary Science