Modern computer vision algorithms allow researchers and art historians to search for artist-characteristic contour extraction from sketches, thus providing accurate input for artwork analysis, for possible assignments and classifications, and also for the identification of the specific stylistic features. We approach this challenging task with three machine learning algorithms and evaluate their performance on a small collection of images from five distinct artists. These algorithms aim to find the most appropriate artist for a sketch (or a contour of a sketch), with promising results that have a higher level of confidence (around 92%). Models start from common Faster R-CNN architectures, reinforcement learning, and vector extraction tools. The proposed tool provides a base for future improvements to create a tool that aids artwork evaluators.

This study explores the use of Temporal Fusion Transformers (TFTs), an AI/ML technique, to enhance the prediction of coastal dynamics along the Western Black Sea coast. We integrate in-situ observations from five meteo-oceanographic stations with modelled geospatial marine data from the Copernicus Marine Service. TFTs are employed to refine predictions of shallow water dynamics by considering atmospheric influences, with a particular focus on wave-wind correlations in coastal regions. Atmospheric pressure and temperature are treated as latitude-dependent constants, with specific investigations into extreme events like freezing and solar radiation-induced turbulence. Explainable AI (XAI) is exploited to ensure transparent model interpretations and identify key influential input variables. Data attribution strategies address missing data concerns, while ensemble modelling enhances overall prediction robustness. The models demonstrate a significant improvement in prediction accuracy compared to traditional methods. This research provides a deeper understanding of atmosphere-marine interactions and demonstrates the efficacy of Artificial intelligence (AI)/Machine Learning (ML) in bridging observational and modelled data gaps for informed coastal zone management decisions, essential for maritime safety and coastal management along the Western Black Sea coast.

Microplastic pollution presents a significant and rising risk to both ecological integrity and the long-term viability of economic activities reliant on marine ecosystems. The Black Sea, a region sustaining economic sectors such as fisheries, tourism, and maritime transport, is increasingly vulnerable to this form of contamination. Mytilus galloprovincialis, a well-established bioindicator, accumulates microplastics, providing a direct measure of environmental pollution and indicating potential economic consequences deriving from degraded ecosystem services. While previous studies have documented microplastic pollution in the Black Sea, our paper specifically quantified microplastic contamination in M. galloprovincialis collected from four sites along the western Black Sea coast, each characterised by distinct levels of anthropogenic influence: Midia Port, Constanta Port, Mangalia Port, and 2 Mai. We used statistical analysis to quantify site-specific microplastic contamination in M. galloprovincialis and employed machine learning to develop models predicting accumulation patterns based on environmental variables. Our findings demonstrate the efficacy of mussels as bioindicators of marine plastic pollution and highlight the utility of machine learning in developing effective predictive tools for monitoring and managing marine litter contamination in marine environments, thereby contributing to sustainable economic practices.

This paper presents a deep learning-based system designed for generating, storing, and retrieving embeddings, specifically tailored for analyzing craquelure networks in paintings. Craquelure, the fine pattern of the craquelure network formed on a painting’s surface over time, is a unique “fingerprint” for artwork item authentication. The system utilizes a modified VGG19 backbone, which effectively balances computational efficiency with the ability to extract rich, multi-scale features from high-resolution grayscale images. By leveraging this architecture, the model captures global structural patterns and local texture information, which are essential for reliable analysis.

The Black Sea, a unique semi-enclosed marine ecosystem, is the eastern maritime boundary of the European Union and holds significant ecological importance. The present study investigates anthropogenic noise pollution in the context of the Marine Strategy Framework Directive’s Descriptor 11, with a particular emphasis on the criteria for impulsive sound (D11C1) and continuous low-frequency sound (D11C2) in Romanian ports, which handle a substantial share of regional cargo traffic, and impact maritime activities and associated noise levels. The noise levels from shipping activity vary across Romanian waters, including territorial waters, the contiguous zone, and the Exclusive Economic Zone. These areas are classified by high, medium, and low ship traffic density. Ambient noise levels at frequencies of 63 Hz and 125 Hz, dominated by shipping noise, were established, along with their hydrospatial distribution for the 2019–2020 period. Furthermore, predictive modeling techniques are used in this study to assess underwater noise pollution from human sources. This modeling effort represents the first initiative in the region and utilizes the BELLHOP ray-tracing method for underwater acoustic channel modeling in shallow-water environments. The model incorporates realistic bathymetry, oceanography, and geology features for environmental input, allowing for improved prediction of acoustic variability due to time-varying sea variations in shallow waters. The study’s findings have important implications for understanding and mitigating anthropogenic noise pollution’s impact on the Black Sea marine ecosystem.

The paper investigates the potential of Artificial Intelligence (AI) and Machine Learning (ML) techniques, specifically Temporal Fusion Transformers (TFTs), to enhance the prediction of coastal dynamics along the Western Black Sea coast. We aim to bridge the gap between in-situ observations from five meteo-oceanographic stations and modelled geospatial marine data from the Copernicus Marine Service. TFTs are employed to refine predictions of shallow water dynamics by considering atmospheric influences, focusing on wave-wind correlations. Atmospheric pressure and temperature are treated as latitude-dependent constants, with specific investigations into extreme events like freezing and solar radiation-induced turbulence. The analysis utilizes a dataset of meteorological information collected by the Maritime Hydrographic Directorate (MHD) since 2015. The study relies on data gathered from seven automated weather stations at lighthouses along the Romanian coastline. The stations, part of the Romanian Navy - Marine Meteorological Surveillance Network, continuously gather meteorological parameters at specific ground-level heights, including wind speed and direction. The Copernicus Marine Service (CMEMS) wave reanalysis dataset for the Black Sea provides a comprehensive record of wave conditions with a spatial resolution of approximately 2.5 km and hourly temporal resolution. Explainable AI (XAI) is exploited to ensure transparent model interpretations and identify key influential input variables, including static, encoder, and decoder variables. Data attribution strategies address missing data concerns, while ensemble modelling enhances overall prediction robustness. The models demonstrate a significant improvement in prediction accuracy compared to traditional methods. This research provides a deeper understanding of atmosphere-marine interactions and demonstrates the efficacy of AI/ML in bridging observational and modelled data gaps for maritime safety and coastal management along the Western Black Sea coast. Acknowledgements: The research of the M.E.M., P.P., M.G.I., and L.D. was conducted as part of the "Forecasting and observing the open-to-coastal ocean for Copernicus users" FOCCUS Project (https://foccus-project.eu/), funded by the European Union (Grant Agreement No. 101133911). Views and opinions expressed are, however, those of the author(s) only and do not necessarily reflect those of the European Union or the European Health and Digital Executive Agency (HaDEA). Neither the European Union nor the granting authority can be held responsible for them. The presented results of the M.E.M., P.P., M.G.I., and L.D. have been carried out with financial support from the Sectorial Research-Development Plan of the Romanian Ministry of National Defence, PSCD 2021–2024 Project (097/2021, 092/2022, 097/2023, 097/2024): „Development of an integrated monitoring system to increase the quality of hydro-oceanographic data in the area of responsibility of the Romanian Naval Forces".Thanks are extended to the relevant departments of INOE-2000 for their help through the "Core Program with the National Research Development and Innovation Plan 2022-2027" with the support of MCID, project no. PN 23 05/2023, contract 11N/2023.