In the Human-Machine Interactions (HMI) landscape, understanding user emotions is pivotal for elevating user experiences. This paper explores Facial Expression Recognition (FER) within HMI, employing a distinctive multimodal approach that integrates visual and auditory information. Recognizing the dynamic nature of HMI, where situations evolve, this study emphasizes continuous emotion analysis. This work assesses various fusion strategies that involve the addition to the main network of different architectures, such as autoencoders (AE) or an Embracement module, to combine the information of multiple biometric cues. In addition to the multimodal approach, this paper introduces a new architecture that prioritizes temporal dynamics by incorporating Long Short-Term Memory (LSTM) networks. The final proposal, which integrates different multimodal approaches with the temporal focus capabilities of the LSTM architecture, was tested across three public datasets: RAVDESS, SAVEE, and CREMA-D. It showcased state-of-the-art accuracy of 88.11%, 86.75%, and 80.27%, respectively, and outperformed other existing approaches.

Vegetation maps are an essential tool for territorial planning, enabling the identification of areas requiring protection and facilitating the study of key ecosystem dynamics such as their evolution over time and the threats they face. These aspects are especially critical in island territories, where their fragmented nature and isolation from the mainland pose significant challenges to the development of such documentation. Traditionally, these maps have relied on local experts, requiring extensive fieldwork, significant time and financial resources. To address these challenges, a novel dataset focused on Gran Canaria (Canary Islands, Spain) is presented, designed to allow researchers to develop and test deep learning models that automatically generate vegetation maps using computer vision techniques. This dataset is unique in the field of aerial image-based semantic segmentation, as it provides detailed annotations for 20 well-defined vegetation communities, going beyond the broad classifications commonly found in existing datasets (e.g., forests or grasslands). Additionally, an alternative version of the dataset includes five non-vegetal classes, such as water bodies, roads, or buildings to support more visually comprehensive segmentation tasks.

The detection and monitoring of asbestos–cement roofing remain a critical public health and environmental challenge, especially in urban and suburban areas where asbestos-containing materials are still widespread due to their extensive use in the 20th century. Although hyperspectral and high-resolution multispectral remote sensing have proven effective for mapping asbestos–cement roofs, many existing approaches rely on proprietary software, limiting transparency, reproducibility, and large-scale adoption. This study presents a fully reproducible, cost-free Python-based workflow for the detection and temporal monitoring of asbestos–cement roofing using high-resolution multispectral WorldView-3 imagery. The workflow integrates atmospheric correction (using the Py6S radiative transfer model), spatial preprocessing, supervised pixel-based classification, postprocessing, and building-level aggregation within an open framework. A Maximum Likelihood Classifier is applied to VNIR and SWIR data using empirically defined roof typologies to enhance class separability. Pixel-level results are aggregated to the building scale through adaptive thresholding enabling the translation of spectral classifications into meaningful building-level information. Tested over the city of Mantua (Italy), the approach achieved reliable classification performance and enabled multi-temporal comparison to identify changes potentially due to roof remediation. Evaluation metrics (precision, recall, and F1-score) highlight the importance of carefully choosing the building-level threshold. By relying exclusively on open-source tools, the workflow enhances transparency, reproducibility, and scalability for long-term monitoring.

Microplastics represent an emerging threat to aquatic ecosystems, human health, and coastal aesthetics, with increasing concern about their accumulation on beaches due to ocean currents, wave action, and accidental spills. Despite their environmental impact, current methods for detecting and quantifying microplastics remain largely manual, time-consuming, and spatially limited. In this study, we propose a deep learning-based approach for the semantic segmentation of microplastics on sandy beaches, enabling pixel-level localization of small particles under real-world conditions. Twelve segmentation models were evaluated, including U-Net and its variants (Attention U-Net, ResUNet), as well as state-of-the-art architectures such as LinkNet, PAN, PSPNet, and YOLOv11 with segmentation heads. Models were trained and tested on augmented data patches, and their performance was assessed using Intersection over Union (IoU) and Dice coefficient metrics. LinkNet achieved the best performance with a Dice coefficient of 80% and an IoU of 72.6% on the test set, showing superior capability in segmenting microplastics even in the presence of visual clutter such as debris or sand variation. Qualitative results support the quantitative findings, highlighting the robustness of the model in complex scenes.

Aims: Vegetation mapping remains a slow and costly process, as traditional approaches rely heavily on expert fieldwork. Therefore, the objective of this study was to develop an efficient and scalable methodology for generating vegetation maps by leveraging computer vision techniques on aerial imagery. Location: The study focused on the island of Gran Canaria, an ecologically rich territory with heterogeneous environments where vegetation mapping is essential for environmental conservation, ecosystem monitoring, and biodiversity assessment. Methods: Deep semantic segmentation techniques were employed to delineate vegetation communities from high‐resolution aerial imagery. A comparative analysis was conducted between widely used segmentation architectures. The methodology incorporated transfer learning with various backbones and evaluated performance across two versions of the dataset: One focused exclusively on vegetation communities and another that additionally included some non‐ecological classes such as shadows, roads, and water bodies. Finally, an aggregation of the vegetation communities was performed based on biological similarity. Results: The results obtained revealed clear performance differences between models, with Feature Pyramid Network (FPN) consistently achieving the highest Dice and IoU scores across all dataset configurations, reaching approximately 70% Dice and 59% IoU in both aggregated versions of the dataset. The analysis also highlighted the benefits of class aggregation for improving segmentation quality in highly fragmented vegetation types. A final discussion examined these findings and outlined the methodological limitations and practical implications for ecological mapping. Conclusions: The findings confirmed that deep learning‐based semantic segmentation enables the efficient generation of vegetation maps, even in ecologically complex territories. Although performance remained constrained by data availability and class complexity, the results demonstrated that these models can provide accurate and biologically meaningful representations of plant communities. The proposed framework therefore offers a solid foundation for supporting large‐scale ecological monitoring and for guiding future developments toward more detailed, scalable, and data‐rich vegetation mapping strategies.