Search Results (144)

Convolutional neural networks (CNN) have transformed visual recognition, yet robust geometric reasoning, reliable out-of-distribution generalization, and recognition from limited data remain substantially unsolved. CNNs draw their architectural inspiration from the mammalian visual cortex, but the translation from biology to engineering was selective and, in places, imprecise, and those imprecisions have consequences that are well documented. This paper examines where the biological fidelity holds and where it gives way, grounding the analysis in formal results that predate deep learning and in recent empirical findings on CNN failure modes. We identify three diagnosable architectural limitations. First, CNNs conflate visual modalities that the biological system separates structurally at the lateral geniculate nucleus, feeding raw RGB pixels into a single undifferentiated filter bank and entangling orientation, color, and texture signals from the first layer onward. Second, CNNs repeat a spatial subsampling operation across the full depth of the network, far beyond the early visual cortex stages where it has biological warrant. Barnard and Casasent established formally in 1990 that this operation discards positional information irreversibly at every layer where it is applied, and repeating it into regions that correspond to V4 and inferotemporal cortex compounds this loss without the compensating transition to qualitatively different computations that the biological hierarchy performs. Third, the pooling-as-complex-cell analogy that motivated this design reflects a misreading of what complex cells compute. The spatiotemporal energy model formalizes complex cell behavior as geometry extraction: detecting the presence and orientation of a local edge structure robustly, abstracting over photometric accidents of contrast polarity and sub-wavelength phase that are not geometrically meaningful. Pooling is a tolerable first-stage approximation of this behavior, but as a general-purpose invariance mechanism repeated across the full depth of the network, it is attempting something categorically different, namely object-level position invariance through spatial subsampling, which achieves its goal by discarding exactly the geometric information that the energy model preserves. Treating pooling as a scalable, indefinitely repeatable implementation of complex cell behavior—rather than as a first-stage approximation with a natural biological endpoint at V3—conflates two operations that differ not in degree but in kind, and crucially it removed the principled criterion for confining the S-C operation to early visual cortex: because pooling was understood as a general-purpose invariance mechanism, the field had no architectural reason to stop repeating it. We survey how capsule networks, group-equivariant CNNs, PDE-based networks, and vision transformers each address one or two of these limitations while leaving the others intact. We propose six desiderata that a more biologically complete architecture would need to satisfy and argue that satisfying them requires treating the visual cortex’s solution as a coherent package in which each component depends on the others working correctly, rather than as a menu of independently selectable principles. Full article

Background: The rapid growth of e-commerce has intensified the tension between customer expectations for fast, convenient delivery and the need for more sustainable last-mile logistics. While existing studies have examined speed, price, sustainability, and convenience as separate delivery attributes, less attention has been given to how these dimensions are combined and presented in consumer-facing delivery options. Methods: This study adopts a mixed-methods approach, combining a systematic literature review with structured analysis of publicly available delivery offers on websites across the UK retail and logistics sectors. Results: The findings show that delivery design remains strongly shaped by speed, price visibility, and convenience, while sustainability signals are rarely embedded at the point of customer choice. Although the literature highlights growing interest in green logistics, observed delivery menus suggest a persistent gap between sustainability commitments and their implementation at checkout. Five delivery strategy archetypes are identified, illustrating how firms configure trade-offs among fast delivery, affordability, sustainability signalling, and convenience. Conclusions: The study contributes a four-pillar choice architecture framework for understanding online delivery design. It highlights the need for clearer sustainability communication, greener default options, and stronger alignment among firm strategy, consumer decision-making, and policy support in last-mile delivery. Full article

We present FLIC, a real-world annotated dataset designed for the visual estimation of food leftovers in canteens and other collective catering environments using standard 2D RGB imagery. Collected over 22 days in an operational university canteen, the dataset includes 401 paired image acquisitions of full and leftover trays, each associated with pixel-precise semantic segmentation masks and physically measured food mass. The goal is to support research on the estimation of leftover food mass from tray images, a task that has received limited attention compared to pre-consumption food recognition, despite its relevance for sustainability and operational decision making in food services. Unlike existing food datasets, FLIC jointly provides paired before–after visual observations and reliable mass ground truth, enabling quantitative analysis of food leftovers under realistic conditions without relying on depth or multi-view information. To demonstrate the dataset’s applicability, we rely on the concept of digital density, relating pixel area to food mass, and implement a lightweight, interpretable baseline mass estimation pipeline. This includes an automatic food/no-food segmentation stage, evaluated across multiple deep learning models (U-Net, DABNet, DINOv2+FeatUp, and SAM), followed by an assisted food recognition stage that leverages the fixed daily menu to map broad user input (e.g., “first course” vs. “second course”) to a specific food class. Experimental results highlight both the potential and the intrinsic challenges of visual food leftover estimation. Full article

Pinyon–juniper (PJ) woodlands, one of the most extensive mature and old-growth woodland types in the Western United States, provide critical ecological, cultural, and economic benefits but face increasing threats from climate change, altered disturbance regimes, invasive species, and pests. We developed the PJ Woodland Climate Adaptation Management Menu, a decision support tool designed to guide adaptive, climate-informed management of PJ ecosystems, particularly within the Colorado Plateau ecoregion. The menu was created through an iterative, collaborative process involving literature review, integration of strategies from existing adaptation frameworks, and extensive input from scientists, land managers, and community partners during workshops and focus groups. The menu links specific, evidence-based approaches to each of six broad strategies, including soliciting community input, mitigating disturbance, enhancing and maintaining biodiversity, conserving ecotones, timing actions for optimal outcomes, and accepting climate-driven changes when appropriate. It is intended for use with the Adaptation Workbook to help managers connect local goals and climate vulnerabilities to tailored management tactics. Hypothetical scenarios demonstrate the menu’s application to contrasting PJ woodland conditions, from die-off events to old-growth maintenance. Lessons learned during development underscore the value of early stakeholder engagement, cross-sector collaboration, and balancing diverse ecological objectives. This menu offers a flexible, transferable framework to strengthen climate resilience in PJ woodlands and serves as a model that could improve adaptation planning in other dryland forest ecosystems. Full article

Purpose: This study provides a critical examination of the literature on applying artificial intelligence (AI) and Extended Reality (XR) in restaurant settings and related foodservice operations. It focuses on how AI and XE influence consumer nutrition awareness and decision-making about food choices, and their implications for customer satisfaction, loyalty, and service delivery in foodservice environments. Design/methodology/approach: The study adopts a systematic literature review (SLR) approach following the PRISMA method. An initial search identified over 3900 academic papers published between 2016 and 2025. Studies were selected on the basis of predetermined inclusion and exclusion criteria, and 26 peer-reviewed articles were analyzed. The review provides a conceptual synthesis and develops propositions for practical applications and future research directions. Findings: The review reveals a shift from static systems that rely on optimization, toward adaptive and user-centered solutions that are behavior-oriented. AI applications predominate in the case of calorie tracking, personalized recommendations, and menu planning. Though deployment of XR technologies (e.g., AR and VR) is less prevalent, they offer potential for immersive, and real-time interventions. A key distinction emerges between studies demonstrating empirical effectiveness (e.g., improved understanding and healthier choices) and those focused on technical and/or conceptual developments. To date, there has been limited validation of behavioral impacts in foodservice settings. Originality: This study offers a theory-informed conceptualization of AI and XR applications in restaurant and foodservice contexts by integrating three perspectives: hospitality (menus and dining experience), nutrition (dietary awareness and healthier choices), and human–technology interaction (technology acceptance and user engagement). The study reconceptualizes AI- and XR-enabled systems as behavioral intervention tools and outlines a focused research agenda for advancing nutritional communication in foodservice environments. Full article

Virtual reality (VR) interfaces increasingly rely on interaction within peripersonal space. However, the conditions under which interaction performance in virtual environments can approximate those of comparable real-world tasks remain underexplored, particularly for hierarchical menus requiring precise sequential input. This study investigated how the presence or absence of tactile feedback influences movement time and selection accuracy during hierarchical menu interaction in peripersonal space across different task difficulty levels. Twelve participants performed a three-level hierarchical selection task on a 4 × 3 menu in two controlled experiments with a stereoscopic 3D TV. Two interaction conditions were tested: a surface-based condition, with the menu attached to the physical screen, and a mid-air condition, with the menu positioned 35 cm and 45 cm in front of participants. Selections were confirmed using a handheld remote. Results showed no statistically significant difference in movement time and selection accuracy between the virtual and real environments when screen-surface targets provided tactile feedback, but performance declined for mid-air targets without tactile references, particularly under higher task difficulty levels. These findings suggest that tactile feedback, coordinated visual target placement, and users’ familiarity with touchscreen-like interaction jointly act as key factors for designing effective, immersive, and user-friendly VR menu systems in peripersonal space. Full article

Background/Objectives: Metabolic dysfunction-associated steatotic liver disease (MASLD) is a common chronic liver disorder linked to obesity, insulin resistance, and dyslipidemia. Nutrition plays a central role in modulating hepatic lipid metabolism, oxidative stress, and inflammation, yet practical, evidence-based dietary strategies remain limited. This study aimed to develop Mediterranean diet-based meal plans with varying macronutrient compositions and to characterize their nutritional profiles, as well as to evaluate them using established nutritional indices and diet score calculations, such as the Dietary Inflammatory Index, Dietary Antioxidant Index, and dietary lipid indices. Methods: Clinical practice guidelines (CPGs) from various academic and professional organizations were reviewed to assess current non-pharmacological treatments for MASLD, with a focus on determining whether the Mediterranean diet is the most recommended dietary pattern. Traditional, low-carbohydrate, and low-fat MedDiet patterns were translated into food-based meal plans. A 7-day meal plan was developed and analyzed for nutrient composition, then evaluated using the Dietary Inflammatory Index (DII), Dietary Antioxidant Index (DAI), Dietary Lipophilic Index (DLI), and Dietary Lipophilic Load (DLL). A Western diet (WD) that is characterized by ultra-processed food (UPF) was included as a comparative reference. Results: The validated dietary score calculations showed that all MedDiet patterns demonstrated consistently low DII scores (−2.00 to −2.81) and high DAI scores (3 to 20.03), whereas the WD showed high DII scores (5.0 to 6.09) and low DAI scores (−12.47 to −17.99). Despite these variations in macronutrients, the menu developed in the study on three MedDiet patterns showed negative DII and positive DAI scores. When comparing the traditional MedDiet with the WD, which have similar macronutrient distributions, the WD was characterized by less favorable DII and DAI scores. Conclusions: This study provides a descriptive, guideline-informed framework for Mediterranean diet-based meal plans with varying macronutrient compositions. Utilizing DII, DAI, DLI, and DLL offers a potential framework for designing dietary interventions. Further validation through clinical studies is needed to justify the potential for practical and digital translation. Nevertheless, the study provides initial insights that may inform future research on nutritional approaches for MASLD integrating dietary indices. Full article

As a proactive response, reserving a certain amount of relief materials in advance is crucial for responding to potential disasters. Different from public tendering and bidding, this study proposes the purchasing mode of authorization, under which a nonprofit organization (NPO), as a buyer, wholly authorizes the procurement of relief materials to a professional agent. The relief material procurement system under the purchasing mode of authorization is regarded as a bilevel relief supply chain consisting of one buyer, one agent, and two suppliers with private information about the quality levels of relief materials. For the disclosure of private information, the quality-related procurement strategy is designed in the form of a menu based on the suppliers’ private information. A bilevel optimization model is developed based on agency theory to derive the optimal strategic decisions, and the impacts of the main influencing factors on the optimal procurement strategy and the buyer’s minimum expected cost are discussed via numerical analysis. Then, the study is extended by exploring supplier’s alternative cost functions and supply availability, as well as proposing future research directions. This paper presents an optimal quality-related procurement strategy, which provides rules for quickly responding to the changes in influencing factors during the material procurement process, as well as the minimum expected cost for the buyer to purchase relief materials, which serves as a threshold for screening a reliable retail enterprise as the agent. Finally, three managerial implications with practical significance, drawn from our findings, are presented to facilitate cooperation between NPO and large retail enterprises in order to achieve effective procurement of relief materials at the pre-disaster preparation stage. Full article

This study examines how regional gastronomic identity is expressed through restaurant menus in the Okanagan Valley, designated in October 2025 as Canada’s first UNESCO City of Gastronomy. This article aims to assess the emergence of a regional gastronomic identity and the potential recognition of a signature dish. An exploratory sequential mixed-methods approach was used to collect data from 40 restaurants, where 283 main dishes were selected and analyzed. These data were coded primarily to identify recurring compositional structures and emerging ingredient patterns. Several recurring compositional templates appear across restaurants, structured around shared protein–starch–sauce configurations. Although they remain occasional, their repetition across restaurants points to the early formation of a recognizable gastronomic identity. This identity does not rely on a single signature dish but takes shape through shared dish structures that recur across menus. These patterns contribute to ongoing discussions in gastronomy tourism by showing how regional identity can develop through distributed and processual culinary practices. The study shows how menu analysis provides a valuable lens for understanding the development of such an identity in emerging gastronomic destinations. Full article

Clinical diagnostic laboratories continue to face growing pressure from rising test volumes, increasingly complex testing menus, significant workforce shortages, and expectations for faster turnaround times at sustainable cost. Total laboratory automation (TLA) has become a central strategy for improving efficiency in high-volume laboratories, where integrated systems from Abbott, Roche, Siemens Healthineers, and Beckman Coulter have demonstrated substantial reductions in turnaround time, error rates, and labor requirements. Evidence across multiple health systems shows that TLA improves performance and stabilizes laboratory operations even during workload peaks. Despite these gains, large segments of pre-analytical and post-analytical workflows remain manual, especially tasks related to specimen transportation, bench-level manipulation, instrument tending, and troubleshooting. Recent progress in collaborative robotics (cobots), autonomous mobile robots (AMRs), and hospital service robots demonstrates that these technologies can complement TLA by addressing not only the logistical and dexterous tasks that fixed automation lines cannot reach but also enabling robots that can work safely right alongside humans in a shared space. Cobots have shown sub-millimeter precision in colony picking and other fine-motor tasks, though typically at lower throughputs than dedicated track modules, and AMRs have demonstrated reliable transport of pathology carts and medical supplies through large clinical environments. Meanwhile, humanoid-capable mobile manipulators, like Moxi from Diligent Robotics, deployed in hospitals are already completing hundreds of thousands of supply deliveries, indicating real-world significance. Here, we integrate technical, regulatory, operational, and business perspectives on TLA, collaborative robotics, and mobile platforms. We discuss real-world efficiency gains, regulatory expectations under the CLIA and United States FDA, and the emerging case for hybrid automation ecosystems that combine TLA islands, cobotic workcells, AMRs, and AI-enabled orchestration. We argue that the next decade of laboratory automation will move beyond monolithic tracks with robots toward flexible, modular robotic systems designed to operate safely together with humans and to augment the increasingly strained laboratory workforce. This not only allows clinical staff to dedicate more time to patient care but also ensures greater reliability and scalability for essential services throughout demanding hospital environments. Full article

Food systems are major drivers of global environmental change, accounting for about one-third of global greenhouse gas (GHG) emissions and contributing to land degradation, freshwater depletion, and biodiversity loss. Within this system, post-retail activities generate an estimated 18–20% of total food-related GHG emissions. In Europe, food service is responsible for roughly 12% of total food waste, making collective catering a strategic sector for sustainability interventions. Objective: Through menu design and composition, collective catering services can influence the environmental performance of thousands of meals served daily. This study introduces a novel meal-level scoring system—the App for the Environmental Impact Assessment of Dishes in Collective Catering (EcoRistApp, ERA)—designed to assess and communicate the environmental performance of institutional canteen dishes. Methods: EcoRistApp was developed and applied to a representative selection of first courses, second courses, and side dishes. Environmental impacts were quantified using Life Cycle Assessment (LCA) with SimaPro 9.5.5 software and the ReCiPe Midpoint (H) method. Normalized and weighted impact results were aggregated into a composite Environmental Impact Index (EII), which was then translated into a five-color interpretative scale to enhance usability and comprehension. Results: The analysis highlighted marked differences in environmental performance among dishes, largely driven by ingredient type and origin. Plant-based meals, such as lentil soup, consistently achieved lower impact scores, while dishes containing animal-derived ingredients, particularly beef and fish, showed higher impacts across multiple categories. Recipes combining high- and low-impact ingredients demonstrated potential for reducing overall environmental burdens. Conclusions: By converting complex LCA outcomes into an intuitive scoring system, EcoRistApp supports informed decision-making by catering operators and consumers, encourages plant-forward menu strategies, and contributes to the environmental transition of food service systems. Full article

Floods are among the most destructive natural disasters, posing major challenges to human safety, property, and urban resilience. Effective communication of flood risk is therefore crucial for disaster preparedness and the sustainable management of smart cities. This study explores how interface design elements of flood hazard maps, including interaction modes and legend color schemes, influence users’ risk perception, decision support, and usability. An online questionnaire survey (N = 776) and a controlled 2 × 2 experiment (N = 40) were conducted to assess user comprehension, cognitive load, and behavioral responses when interacting with different visualization formats. Results show that slider-based interaction significantly reduces task completion and map-reading times compared with drop-down menus, enhancing usability and information efficiency. Multicolor legends, although requiring higher cognitive effort, improve users’ risk perception, engagement, and memory of flood-related information. These findings suggest that integrating cognitive principles into interactive design can enhance the effectiveness of digital disaster communication tools. By combining human–computer interaction, visual cognition, and smart governance, this study provides evidence-based design strategies for developing intelligent and user-centered flood hazard mapping systems. The proposed framework contributes to the advancement of smart urban resilience and supports the broader goal of building safer and more sustainable cities. Full article

Food recommender systems are pivotal in helping people make optimal dietary choices based on tremendous amounts of data. Extant studies offer different methods and techniques, but the combination of similarity search, large language models (LLMs), and multi-criteria decision-making (MCDM) remains underexplored. This study proposes a new system that leverages all three. First, we utilize an LLM to suggest queries from the same domain as the dish database. Then, the queries are vectorized and used for similarity search to generate a preliminary list of suggested menu items. Next, multiple LLMs provide scores for each item, which become the MCDM inputs, where Lin’s concordance correlation coefficient (LCCC) enhances the weighted sum scalarization technique. We evaluated the prototype on three publicly available dish datasets and at classification thresholds of 0.25, 0.50, and 0.75, and the proposed domain-adaptation approach consistently outperformed the baseline query. For example, at the 0.50 threshold, precision ranged from 49.11% to 56.60%, compared with 35.40% for the baseline. Furthermore, aggregating multiple LLMs mitigates single-model bias in recommendations. To substantiate this, a bootstrap evaluation of the proposed LCCC-based consensus weighting confirms that both the estimated weights and the induced rankings are numerically stable under sampling perturbations. To further ensure the robustness and reliability of the proposed system, we validate the results against other established weighting schemes and state-of-the-art MCDM methods. Moreover, Kendall’s $\tau$-based comparisons across weighting schemes and multiple MCDM methods confirm that the proposed LCCC-based framework produces highly consistent and statistically significant rankings, demonstrating strong robustness to methodological choices. This paper contributes a system architecture and design that can be adopted for other domains of recommender systems where the capability of multiple LLMs can benefit complex and multifaceted decision-making processes. Full article

Background/Objectives: Modern families face challenges in maintaining healthy and sustainable diets due to time constraints and busy lifestyles. The Mediterranean diet (MD), known for its benefits to both personal health and environmental sustainability, is often difficult to apply consistently within households. This paper presents and validates the SWITCHtoHEALTHY AI-based Family Nutrition Recommendation System, designed to generate meal plans aligned with MD guidelines. Methods: Two complementary recommendation engines were developed: the AI-based Family Nutritional Recommender, which creates personalized meal plans for adults that include shared family meals, and the Child Nutritional Recommender, which generates meal plans for children that could also incorporate school menus or proposals from the school cafeteria. Both systems rely on an expert-validated dataset of Mediterranean foods and are designed to comply with the expert-validated nutritional rules based on MD principals and national dietary guidelines. Results: The recommendation systems were validated using data from a real-world family intervention, achieving 90% accuracy in generating meal plans for all family members, while meeting the expert validated dietary rules for both adults and children. Moreover, AI-based Family Nutritional Recommender exceeds 90% accuracy in estimating calorie and nutrient content for adults. Conclusions: The results demonstrate the preliminary potential of AI-based recommendation systems to facilitate healthier and more sustainable dietary habits within modern households by generating personalized, nutritionally balanced family meal plans consistent with MD principles. Full article

Background/Objectives: The university stage is a critical period for consolidating dietary habits that influence future health. University canteens therefore play a key role in providing menus aligned with nutritional recommendations. As menu composition shapes students’ access to healthy food, its evaluation also has equity implications. This study aimed to apply a newly designed questionnaire—the COMUNI questionnaire—intended to provide a rapid, user-friendly, and transferable method for evaluating the dietary quality of lunch menus offered in university canteens. Methods: Two versions of the 13-item COMUNI questionnaire were developed: COMUNI-1 for single-option menus and COMUNI-2 for menus offering multiple first- and second-course choices. The tool evaluates the frequency of key food groups, the availability of water and wholegrain bread, and the variety of foods and culinary techniques. To test the questionnaire, it was applied to 34 menu templates from university residences, colleges, and cafeterias. Results: 85.3% of menus showed deficient dietary quality, and 14.7% were rated as improvable; none achieved an optimal score. Menus managed by catering companies obtained significantly higher scores than those under direct management. Most frequently shortcomings included insufficient offerings of vegetables, legumes, fish, and wholegrain bread, alongside a frequent presence of refined carbohydrate sources and fried or ultra-processed foods. Conclusions: Universities should incorporate adherence to dietary recommendations as a key criterion in food-service procurement. The COMUNI questionnaire provides a simple and operational tool for assessing menu quality, supporting both diagnosis and monitoring of university food-service, once formally validated. Its use may also help identify structural disparities in access to healthy foods across campus settings, supporting more equitable food-service policies. Full article