Over the last decade, Unmanned Aerial Vehicle (UAV) technology has rapidly developed by integrating AI tools and methodologies. UAVs are used across many fields, including wireless communication, logistics, agriculture, construction, security, and exploration. UAV systems are inherently complex and present numerous challenges. This study conducted a systematic literature review to identify the main challenges and the solutions researchers have proposed. A total of 141 research papers were collected and analyzed, with no restriction on publication year. The review identified 10 main challenges affecting AI in UAVs. These include problems with autonomous navigation, real-time decision-making,... More >

Remote Sensing (RS) data streams necessitate classification models capable of assimilating novel information without the computational burden of retraining from scratch. Although Few-Shot Incremental Learning (FSIL) offers a promising paradigm for adaptation using limited samples, existing methodologies often falter due to the high inter-class similarity and complex background clutter inherent in aerial imagery. To address these challenges, we introduce a novel framework, Geospatial Few-Shot Adaptive Compatible Training, designed to mitigate catastrophic forgetting in dynamic environments. This method ensures forward compatibility by strategically allocating embedding space for future catego... More >

Crowdsourced software development (CSD) has gained increasing popularity in software engineering. However, empirical evidence on its sustainability remains limited. Although previous studies have suggested several sustainability-related factors, their applicability in real industrial CSD environments lacks sufficient validation. Through a systematic literature review (SLR), we identified eleven factors affecting sustainability in CSD. This study empirically validates these factors via a questionnaire survey with practitioners from various crowdsourced software development platforms. The results confirm that all eleven factors identified in the literature are also recognized by practitioners,... More >

Accurate liver segmentation from three-dimensional (3D) computed tomography (CT) volumes is a critical step in computer-aided diagnosis, surgical planning, and disease quantification. Despite substantial progress in deep learning, achieving robust and generalizable liver segmentation remains challenging due to complex organ boundaries, pathological variations, and domain shifts across scanners. This review provides a comprehensive overview of 3D volumetric liver segmentation techniques, spanning from classical model-based methods to contemporary transformer-driven frameworks. We categorize existing methods into three paradigms: (1) classical statistical and atlas-based methods, (2) deep conv... More >

Presents the introductory editorial for the inaugural issue of this title. More >