Abstract
On a CPU-FPGA system, C/C++ applications are typically accelerated by offloading specific code regions onto the FPGA using High-level Synthesis (HLS). Although modern FPGAs can implement increasingly large and complex designs, the size and variety of potential offloading code regions remain constrained by the limitations of HLS tools (e.g., no support for dynamic memory allocation and system calls). This paper proposes automated C/C++ source-to-source transformations that tackle these limitations in two steps. Firstly, transformations reduce the entropy of an input C/C++ application by converting it into a subset of C, e.g., by flattening arrays and structs. Secondly, additional transformations make a selected code region synthesizable, e.g., by moving dynamic memory allocations out of the region, converting them to static memory, and offloading non-synthesizable C standard library calls, such as printf(), to the CPU. We evaluate the impact of these transformations showing results obtained through Vitis HLS for four real-world examples: the disparity and texture-synthesis benchmarks from CortexSuite, which contain dynamic memory allocations and indirect pointers in their hotspots; llama2, a Large Language Model that calls printf() every time it predicts a new word; and the spam-filter benchmark from Rosetta, as a debugging showcase. © 2025 IEEE.

Abstract
In geo-replicated distributed systems, data is redundantly stored across nodes at different geographical sites, increasing fault tolerance and ensuring low access latency by placing data closer to the end user. With data being concurrently updated across sites, replicas should converge to a consistent view of the data, which leads toward adopting fine-tuned consistency models, namely causal consistency (CC). On the one hand, CC respects the causality between operations, resulting in intuitive outcomes for end users and programmers. On the other hand, it avoids the latency penalty of stronger consistency models and bypasses their availability constraints in the presence of network partitions. Furthermore, when coupled with read-only transactions (ROTs) capable of extracting a unified view of the data, CC avoids the anomalies of weaker consistency models. ROTs, however, cause additional coordination overhead compared to non-transactional reads. This overhead is particularly unwelcome considering the prevalence of read operations in real-world applications, and hence the impact of ROTs on the overall performance of read-heavy systems. With this in mind, there has been a growing effort to optimize the latency and throughput of causally consistent ROTs and to understand how the design of existing systems impacts their performance. In light of these recent developments, the present work surveys the state-of-the-art of causally consistent distributed systems, summarizing and comparing their core characteristics and tradeoffs and examining how their design decisions impact the performance of ROTs. To this end, it first defines some key concepts and presents two impossibility results concerning the properties of ROT algorithms. It then reviews several causally consistent systems with ROT support by identifying their recurring strategies to ensure causality and summarizing each of their designs and properties, stressing their implications on the performance of ROTs. It also surveys two architectural approaches to CC, which present progress toward a standard implementation for causally consistent systems. Finally, it discusses the open challenges identified in the literature.

Abstract
Virtual Reality applications increasingly require methods to effectively guide users to important elements within the virtual environment. Central visual cues are the most common method, which have proven effective for directing attention, yet often compromise on level of immersion. This work explored whether peripheral visual cues could serve as an alternative approach that supports attention guidance while preserving sense of presence. We performed a user study with 24 participants to compare four visual cues: two central cues (Floating Text and Floating Arrow) and two peripheral cues (Edge Lighting and Swarm). Users completed a visual search task of 7 objects for each visual cue, with data collected on performance through reaction time, round time, and total errors. Additionally, presence and workload were evaluated through the IGROUP Presence Questionnaire and NASA Task Load Index, respectively. No statistically significant differences were found between peripheral and central cues for presence, however performance and workload varied significantly based on specific cue implementation rather than type of positioning. Our findings indicate that peripheral positioning does not inherently provide attention guidance advantages over central placement. Instead, thoughtful cue design, with a simple yet clear appearance and behavior appears to be the critical factor for achieving effective attention guidance while preserving presence in IVEs. These results provide valuable insights for VR content creators to facilitate the design process of VR experiences.

Abstract
Nowadays, eXtended Reality (XR) has matured to the point where it seamlessly integrates various input and output modalities, enhancing the way users interact with digital environments. From traditional controllers and hand tracking to voice commands, eye tracking, and even biometric sensors, XR systems now offer more natural interactions. Similarly, output modalities have expanded beyond visual displays to include haptic feedback, spatial audio, and others, enriching the overall user experience. In this vein, as the field of XR becomes increasingly multimodal, the education process must also evolve to reflect these advancements. There is a growing need to incorporate additional modalities into the curriculum, helping students understand their relevance and practical applications. By exposing students to a diverse range of interaction techniques, they can better assess which modalities are most suitable for different contexts, enabling them to design more effective and human-centered solutions. This work describes an Advanced Human-Machine Interaction (HMI) course aimed at Doctoral Students in Computer Science. The primary objective is to provide students with the necessary knowledge in HMI by enabling them to articulate the fundamental concepts of the field, recognize and analyze the role of human factors, identify modern interaction methods and technologies, apply HCD principles to interactive system design and development, and implement appropriate methods for assessing interaction experiences across advanced HMI topics. In this vein, the course structure, the range of topics covered, assessment strategies, as well as the hardware and infrastructure employed are presented. Additionally, it highlights mini-projects, including flexibility for students to integrate their projects, fostering personalized and project-driven learning. The discussion reflects on the challenges inherent in keeping pace with this rapidly evolving field and emphasizes the importance of adapting to emerging trends. Finally, the paper outlines future directions and potential enhancements for the course.

Abstract
Cardiovascular diseases are the leading cause of death in the world, with coronary artery disease being the most prevalent. Coronary artery calcifications are critical biomarkers for cardiovascular disease, and their quantification via non-contrast computed tomography is a widely accepted and heavily employed technique for risk assessment. Manual segmentation of these calcifications is a time-consuming task, subject to variability. State-of-the-art methods often employ convolutional neural networks for an automated approach. However, there is a lack of studies that perform these segmentations with 3D architectures that can gather important and necessary anatomical context to distinguish the different coronary arteries. This paper proposes a novel and automated approach that uses a lightweight three-dimensional convolutional neural network to perform efficient and accurate segmentations and calcium scoring. Results show that this method achieves Dice score coefficients of 0.93 ± 0.02, 0.93 ± 0.03, 0.84 ± 0.02, 0.63 ± 0.06 and 0.89 ± 0.03 for the foreground, left anterior descending artery (LAD), left circumflex artery (LCX), left main artery (LM) and right coronary artery (RCA) calcifications, respectively, outperforming other state-of-the-art architectures. An external cohort validation also showed the generalization of this method's performance and how it can be applied in different clinical scenarios. In conclusion, the proposed lightweight 3D convolutional neural network demonstrates high efficiency and accuracy, outperforming state-of-the-art methods and showcasing robust generalization potential.

Abstract
Online reviews are a crucial asset for e-commerce platforms as they provide consumers with valuable insights into products. It is important to note that these reviews are subjective and may contain biases. Therefore, it is essential to approach them with a critical eye. Despite this, online reviews remain a valuable tool for consumers when making purchasing decisions. This study focuses on developing web-based mini-games that target cognitive biases. The games are specifically designed to enhance the perception of e-commerce online reviews. A pilot study involving 85 participants was conducted to explore the potential of integrating these cognitive bias games into web platforms. The findings indicate promising avenues for leveraging these games to enhance cognitive personalization and improve the quality of e-commerce online reviews.