Abstract
General-purpose AI-assisted tools, such as ChatGPT, have recently gained much attention from the media and the general public. That raised questions about in which tasks we can apply such a tool. A good code design is essential for agile software development to keep it ready for change. In this context, identifying which design pattern can be appropriate for a given scenario can be considered an advanced skill that requires a high degree of abstraction and a good knowledge of object orientation. This paper aims to perform an exploratory study investigating the effectiveness of an AI-assisted tool in assisting developers in choosing a design pattern to solve design scenarios. To reach this goal, we gathered 56 existing questions used by teachers and public tenders that provide a concrete context and ask which design pattern would be suitable. We submitted these questions to ChatGPT and analyzed the answers. We found that 93% of the questions were answered correctly with a good level of detail, demonstrating the potential of such a tool as a valuable resource to help developers to apply design patterns and make design decisions.

Abstract
Recent state-of-the-art forecasting methods are trained on collections of time series. These methods, often referred to as global models, can capture common patterns in different time series to improve their generalization performance. However, they require large amounts of data that might not be available. Moreover, global models may fail to capture relevant patterns unique to a particular time series. In these cases, data augmentation can be useful to increase the sample size of time series datasets. The main contribution of this work is a novel method for generating univariate time series synthetic samples. Our approach stems from the insight that the observations concerning a particular time series of interest represent only a small fraction of all observations. In this context, we frame the problem of training a forecasting model as an imbalanced learning task. Oversampling strategies are popular approaches used to handle the imbalance problem in machine learning. We use these techniques to create synthetic time series observations and improve the accuracy of forecasting models. We carried out experiments using 7 different databases that contain a total of 5502 univariate time series. We found that the proposed solution outperforms both a global and a local model, thus providing a better trade-off between these two approaches.

Abstract
The SEAGUARD concept addresses a multi-domain (air, sea, underwater) maritime surveillance approach, involving the deployment, management and coordination of heterogeneous platforms, sensors and information technologies. SEAGUARD’s aim is to deliver a high level of situational awareness through a holistic surveillance system fitted to the needs and ambition of modern border management authorities. The operational context of large maritime areas and the nature of threats - increasingly dynamic, transnational and highly mobile - reflect the growing need to have multiple and different types of authorities involved in and coordinating response efforts so that, working together, their common goals are achieved, with superior efficiency and effectiveness. Attaining the SEAGUARD vision requires a high level of interoperability between the diverse and heterogeneous participating entities (organizations, units, people) and technological systems (unmanned platforms and smart devices) in a collective working in a civil- military, cross-organisation and cross-border environment. To enable this advanced synchronization, the SEAGUARD Interoperability Framework (S.IF) implements a set of Command and Control (C2) rules and protocols among participating entities, benefitting from NATO C2 Approaches as foundational references for its novel interoperability approach.

Abstract
Large-scale image datasets are often partially labeled, where only a few categories' labels are known for each image. Assigning pseudo-labels to unknown labels to gain additional training signals has become prevalent for training deep classification models. However, some pseudo-labels are inevitably incorrect, leading to a notable decline in the model classification performance. In this paper, we propose a new method called Category-wise Fine-Tuning (CFT), aiming to reduce model inaccuracies caused by the wrong pseudo-labels. In particular, CFT employs known labels without pseudo-labels to fine-tune the logistic regressions of trained models individually to calibrate each category's model predictions. Genetic Algorithm, seldom used for training deep models, is also utilized in CFT to maximize the classification performance directly. CFT is applied to well-trained models, unlike most existing methods that train models from scratch. Hence, CFT is general and compatible with models trained with different methods and schemes, as demonstrated through extensive experiments. CFT requires only a few seconds for each category for calibration with consumer-grade GPUs. We achieve state-of-the-art results on three benchmarking datasets, including the CheXpert chest X-ray competition dataset (ensemble mAUC 93.33%, single model 91.82%), partially labeled MS-COCO (average mAP 83.69%), and Open Image V3 (mAP 85.31%), outperforming the previous bests by 0.28%, 2.21%, 2.50%, and 0.91%, respectively. The single model on CheXpert has been officially evaluated by the competition server, endorsing the correctness of the result. The outstanding results and generalizability indicate that CFT could be substantial and prevalent for classification model development. Code is available at: https://github.com/maxium0526/category-wise-fine-tuning.

Abstract
This work introduces Dialogue-AV, a benchmarking dataset for Audio-Video-Language (AVL). We propose using dialogue to describe video content instead of single captions, capturing nuances and shared meanings between audio and visual elements. This approach contributes significantly to improving the diversity of video descriptions and enables comprehensive evaluation of AVL learning across different downstream tasks, such as Cross-Modal Retrieval, Visual Question-Answering, and Video Captioning. Our dataset comprises approximately 258k audiovisual samples accompanied by dialogue-based descriptions for benchmarking. Dialogue-AV builds upon existing State-of-the-Art (SOTA) datasets that feature human-generated descriptions, enhancing them with model-generated ones that describe all modalities. We also present zero-shot baseline results utilising SOTA Visual-Language Models (VLMs), demonstrating that Dialogue-AV is capable of benchmarking a variety of downstream tasks with diverse inputs. Our key contributions include: 1) Dialogue-AV, a benchmark dataset for dialogue-based AVL models; and 2) benchmarks that expose the limitations of current SOTA VLMs. The code and dataset are accessible at: github.com/lvilaca16/dialogue-av. © 2025 IEEE.

Abstract
Advanced rider assistance systems (ARAS) play a crucial role in enhancing motorcycle safety through features such as collision avoidance, blind-spot detection, and adaptive cruise control, which rely heavily on sensors like radar, cameras, and LiDAR. However, their performance is often compromised under adverse weather conditions, leading to sensor interference, reduced visibility, and inconsistent reliability. This study evaluates the effectiveness and limitations of ARAS technologies in rain, fog, and snow, focusing on how sensor performance, algorithms, techniques, and dataset suitability influence system reliability. A thematic analysis was conducted, selecting studies focused on ARAS in adverse weather conditions based on specific selection criteria. The analysis shows that while ARAS offers substantial safety benefits, its accuracy declines in challenging environments. Existing datasets, algorithms, and techniques were reviewed to identify the most effective options for ARAS applications. However, more comprehensive weather-resilient datasets and adaptive multi-sensor fusion approaches are still needed. Advancing in these areas will be critical to improving the robustness of ARAS and ensuring safer riding experiences across diverse environmental conditions.