Search Results (403)

High-resolution range profile (HRRP) target recognition has garnered significant attention in radar automatic target recognition (RATR) research for its rich structural information and low computational costs. With the rapid advancements in deep learning, methods for HRRP target recognition that leverage deep neural networks have emerged as the dominant approaches. Nevertheless, these traditional closed-set recognition methods require labeled data for every class in training, while in reality, seen classes and unseen classes coexist. Therefore, it is necessary to explore methods that can identify both seen and unseen classes simultaneously. To this end, a semantic- and physical-guided generative network (SPGGN) was innovatively proposed for HRRP generalized zero-shot recognition; it combines a constructed knowledge graph with attribute vectors to comprehensively represent semantics and reconstructs strong scattering points to introduce physical constraints. Specifically, to boost the robustness, we reconstructed the strong scattering points from deep features of HRRPs, where class-aware contrastive learning in the middle layer effectively mitigates the influence of target-aspect variations. In the classification stage, discriminative features are produced through attention-based feature fusion to capture multi-faceted information, while the design of balancing loss abates the bias towards seen classes. Experiments on two measured aircraft HRRP datasets validated the superior recognition performance of our method. Full article

Background: Steel box girders are widely employed in bridge engineering due to their excellent mechanical properties and construction convenience, yet their modular design still encounters bottlenecks such as knowledge reuse difficulties and information silos. This study proposes a BIM-driven framework based on knowledge graphs and data fusion. By constructing a professional knowledge graph comprising 85 core entity types and 150 semantic relationships (integrated with over 15,000 knowledge units), systematic management of design knowledge is achieved. The developed BIM reverse modeling technology improves parametric modeling efficiency by 30–40%, while the data fusion mechanism supports over 90% accuracy in design conflict detection. The intelligent decision-making system built upon this framework meets 75% of business scenario requirements while effectively assisting critical decisions such as module selection. Results demonstrate that this framework significantly enhances design collaboration efficiency and intelligence through knowledge structuring and deep data integration. Although some achievements were validated via simulation due to limited field measurement data, the approach demonstrates strong engineering applicability and provides novel technical pathways and methodological support for advancing digital transformation in bridge engineering. Full article

This paper explores the capability of Large Language Models (LLMs) to perform zero-shot planning through multimodal reasoning, with a particular emphasis on applications to Unmanned Ground Vehicles (UGVs) and unmanned platforms in general. We present a modular system architecture that integrates a general-purpose LLM with visual and spatial inputs for adaptive planning to iteratively guide UGV behavior. Although the framework is demonstrated in a ground-based setting, it directly extends to other unmanned systems, where semantic reasoning and adaptive planning are increasingly critical for autonomous mission execution. To assess performance, we employ a continuous evaluation metric that jointly considers distance and orientation, offering a more informative and fine-grained alternative to binary success measures. We evaluate a foundational LLM (i.e., Gemini 2.0 Flash, Google DeepMind) on a suite of zero-shot navigation and exploration tasks in simulated environments. Unlike prior LLM-robot systems that rely on fine-tuning or learned waypoint policies, we evaluate a purely zero-shot, stepwise LLM planner that receives no task demonstrations and reasons only from the sensed data. Our findings show that LLMs exhibit encouraging signs of goal-directed spatial planning and partial task completion, even in a zero-shot setting. However, inconsistencies in plan generation across models highlight the need for task-specific adaptation or fine-tuning. These findings highlight the potential of LLM-based multimodal reasoning to enhance autonomy in UGV and drone navigation, bridging high-level semantic understanding with robust spatial planning. Full article

Existing salient object detection methods for optical remote sensing images still face certain limitations due to complex background variations, significant scale discrepancies among targets, severe background interference, and diverse topological structures. On the one hand, the feature transmission process often neglects the constraints and complementary effects of high-level features on low-level features, leading to insufficient feature interaction and weakened model representation. On the other hand, decoder architectures generally rely on simple cascaded structures, which fail to adequately exploit and utilize contextual information. To address these challenges, this study proposes a Hierarchical Semantic Interaction Module to enhance salient object detection performance in optical remote sensing scenarios. The module introduces foreground content modeling and a hierarchical semantic interaction mechanism within a multi-scale feature space, reinforcing the synergy and complementarity among features at different levels. This effectively highlights multi-scale and multi-type salient regions in complex backgrounds. Extensive experiments on multiple optical remote sensing datasets demonstrate the effectiveness of the proposed method. Specifically, on the EORSSD dataset, our full model integrating both CA and PA modules improves the max F-measure from 0.8826 to 0.9100 (↑2.74%), increases maxE from 0.9603 to 0.9727 (↑1.24%), and enhances the S-measure from 0.9026 to 0.9295 (↑2.69%) compared with the baseline. These results clearly demonstrate the effectiveness of the proposed modules and verify the robustness and strong generalization capability of our method in complex remote sensing scenarios. Full article

The proliferation of Unmanned Aerial Vehicles (UAVs) enables the large-scale collection of ecological data, yet translating this dynamic sensor data into engaging, personalized public experiences remains a significant challenge. Existing solutions fall short: static exhibitions lack adaptability, while general-purpose LLM agents struggle with real-time responsiveness and reliability. To address this, we introduce CurationAgent, a novel intelligent agent built upon the State-Gated Agent Architecture (SGAA). Its core innovation is an advanced hybrid curation pipeline that synergizes Retrieval-Augmented Generation (RAG) for broad semantic recall with an Intent-Driven Curation (IDC) Funnel for precise intent formalization and narrative synthesis. This hybrid model robustly translates user intent into a curated, multi-modal narrative. We validate this framework in a proof-of-concept virtual exhibition of the Lalu Wetland’s biodiversity. Our comprehensive evaluation demonstrates that CurationAgent is significantly more responsive (1512 ms vs. 4301 ms), reliable (95% vs. 57% task success), and precise (85.5% vs. 52.7% query precision) than standard agent architectures. Furthermore, a user study with 27 participants confirmed our system leads to measurably higher user engagement. This work contributes a robust and responsive agent architecture that validates a new paradigm for interactive systems, shifting from passive information retrieval to active, partnered experience curation. Full article

High-throughput phenotypic acquisition and analysis allow us to accurately quantify trait expressions, which is essential for developing intelligent breeding strategies. However, there is still much potential to explore in the field of high-throughput phenotyping for edible fungi. In this study, we developed a portable multi-phenotypic acquisition system for Pleurotus eryngii using RGB and RGB-D cameras. We developed an innovative Unet-based semantic segmentation model by integrating the ASPP structure with the VGG16 architecture. This allows for precise segmentation of the cap, gills and stem of the fruiting body. By leveraging depth images from RGB-D cameras, we can effectively collect phenotypic information about Pleurotus eryngii. By combining K-means clustering with Lab color space thresholds, we are able to achieve more precise automatic classification of Pleurotus eryngii cap colors. Moreover, AlexNet is utilized to classify the shapes of the fruiting bodies. The Aspp-VGGUnet network demonstrates remarkable performance with a mean Intersection over Union (mIoU) of 96.47% and a mean pixel accuracy (mPA) of 98.53%. These results reflect respective improvements of 3.03% and 2.23% compared to the standard Unet model, respectively. The average error in size phenotype measurement is just 0.15 ± 0.03 cm. The accuracy for cap color classification reaches 91.04%, while fruiting body shape classification achieves 97.90%. The proposed multi-phenotype acquisition system reduces the measurement time per sample from an average of 76 s (manual method) to about 2 s, substantially increasing data acquisition throughput and providing robust support for scalable phenotyping workflows in breeding research. Full article

Eye tracking scanpaths encode the temporal sequence and spatial distribution of eye movements, offering insights into visual attention and aesthetic perception. However, analysing scanpaths still requires substantial manual effort and specialised expertise, which limits scalability and constrains objectivity of eye tracking methods. This paper examines whether and how multimodal large language models (MLLMs) can provide objective, expert-level scanpath interpretations. We used GPT-4o as a case study to develop eye tracking scanpath analysis (ETSA) approach which integrates (1) structural information extraction to parse scanpath events, (2) knowledge base of visual-behaviour expertise, and (3) least-to-most and few-shot chain-of-thought prompt engineering to guide reasoning. We conducted two studies to evaluate the reliability and effectiveness of the approach, as well as an ablation analysis to quantify the contribution of the knowledge base and a cross-model evaluation to assess generalisability across different MLLMs. The results of repeated-measures experiment show high semantic similarity of 0.884, moderate feature-level agreement with expert scanpath interpretations (F1 = 0.476) and no significant differences from expert annotations based on the exact McNemar test (p = 0.545). Together with the ablation and cross-model findings, this study contributes a generalisable and reliable pipeline for MLLM-based scanpath interpretation, supporting efficient analysis of complex eye tracking data. Full article

Three-dimensional (3D) reconstruction technology for space targets can provide information support such as target structures and dimensions for space missions including on-orbit services and fault diagnosis, which is crucial for maintaining the normal operation of space assets. Optical devices and ISAR (Inverse Synthetic Aperture Radar) can provide high-resolution two-dimensional (2D) images of space targets, serving as the primary means for space target observation. However, existing 3D imaging methods for space targets exhibit significant limitations: the fusion process of optical observation data and ISAR observation data lacks automation, and factors such as image offset that affect 3D imaging quality are not fully considered. To address these issues, this paper proposes a novel 3D imaging method for space targets utilizing optical-ISAR joint observation. This method first employs semantic segmentation networks to automatically extract target regions from optical and ISAR images. Then, it combines octree-space carving technology for efficient 3D reconstruction and performs correction of target region offset based on projection optimization to achieve high-quality 3D imaging. This method eliminates the need for manual target region extraction, improving the automation level of the algorithm. The application of octree-space carving technology greatly enhances the efficiency of 3D reconstruction. Moreover, by correcting target region offset, the proposed method delivers superior 3D imaging results. Simulation experiments demonstrate that the method exhibits significant superior performance in terms of reconstruction efficiency and imaging quality. Additionally, experiments based on measured data further verify the feasibility and practical application value of the proposed method. Full article

Fake review detection is critical for maintaining trust and ensuring decision reliability across digital marketplaces and IoT-enabled ecosystems. This study presents a zero-shot framework for multi-modal anomaly detection in user reviews, integrating textual and metadata-derived signals through instruction-tuned transformers. The framework integrates three complementary components: language perplexity scoring with FLAN-T5 to capture linguistic irregularities, unsupervised reconstruction via a transformer-based autoencoder to identify structural deviations, and semantic drift analysis to measure contextual misalignment between task-specific and generic embeddings. To enhance applicability in sensor-driven environments, the framework incorporates device-level metadata such as timestamps, product usage logs, and operational signals to enable cross-validation between unstructured text and structured sensor features. A unified anomaly score fusing textual and sensor-informed signals improves robustness under multi-modal detection scenarios, while interpretability is achieved through token-level saliency maps for textual analysis and feature-level attributions for sensor metadata. Experimental evaluations on the Amazon Reviews 2023 dataset, supplemented by metadata-rich sources including the Amazon Review Data 2018 and Historic Amazon Reviews (1996–2014) datasets demonstrate strong zero-shot performance (AUC up to 0.945) and additional few-shot adaptability under limited supervision (AUC > 0.95), maintaining stable precision–recall trade-offs across product domains. The proposed framework provides real-world impact by enabling real-time, multi-modal fake review detection in IoT-driven platforms and smart spaces, supporting consumer trust, automated decision-making, and transparent anomaly detection in sensor-enhanced digital ecosystems. Full article

Background/Objectives: The aging global population has led to a rising prevalence of type 2 diabetes mellitus (T2DM), in which biomedical outcomes alone fail to capture patients’ lived experiences. Patient-Reported Outcome Measures (PROMs) can provide insights into psychological, psychosocial, and quality-of-life (QoL) dimensions, yet their use—particularly among older adults—remains inconsistent. This AI-assisted rapid review aimed to map how PROMs are currently applied in adults with T2DM, with specific attention to studies involving older populations, focusing on their role in assessing well-being, distress, depression, and treatment satisfaction, as well as their implementation in clinical and organizational practice. Methods: A rapid review was conducted using Elicit, an AI tool designed to support systematic evidence synthesis. Studies published between 2015 and 2025 were identified from Semantic Scholar, complemented by manual searches for recent or unindexed papers. Eligibility criteria required inclusion of adults with T2DM and use of validated PROMs in real-world settings. Studies explicitly describing older or elderly subgroups were highlighted separately. After screening 504 records, 167 studies were included. Data extraction covered study design, instruments used, populations, outcomes, and implementation details. Results: The most frequently assessed outcomes were diabetes distress, depression, QoL, treatment satisfaction, and self-efficacy. Common instruments included PAID, DDS, PHQ-9, WHO-5, EQ-5D, SF-36, DTSQ, and GDS. Evidence showed PROMs effectively identified high-risk patients and informed tailored interventions, but integration into routine care remained limited. Barriers included workflow disruption, lack of provider training, heterogeneity of tools, and insufficient cultural validation. Facilitators included brief instruments, digital administration, and linkage with care planning. Conclusions: PROMs are valuable in capturing psychosocial and psychological burdens in adults with T2DM, including but not limited to older populations, but routine implementation is inconsistent. Broader adoption will require digital infrastructure, clinician training, and organizational integration, as well as the development of PROMs that capture experiences with emerging diabetes technologies. Methodologically, this study illustrates the feasibility of AI-assisted rapid reviews to generate timely, evidence-informed syntheses. Full article

This paper introduces a learnability framework for validating large language model (LLM) information extraction without ground-truth annotations. Applied to 20,809 Ottoman legal texts, the framework achieves a Learnability Score of 0.891 through multi-classifier consensus, with external validation confirming substantial agreement across five diverse LLMs ($\kappa$ = 0.785) and human experts ($\kappa$ = 0.786). The approach treats internal consistency as a measurable systemic property, where heterogeneous machine learning models independently rediscover LLM-assigned patterns. Confusion analysis reveals errors concentrate at jurisprudentially meaningful boundaries (e.g., commercial-inheritance: 20.4% of disagreements), demonstrating semantic coherence rather than arbitrary noise. The framework offers practical validation for historical and specialized corpora where traditional annotation is infeasible, processing documents at USD 0.01 each with parallelizable throughput. Validated annotations enable knowledge graph construction with 20,809 document nodes, 7 category nodes, and confusion-weighted semantic proximity edges. This systems-based methodology advances reproducible computational research in domains lacking established benchmarks. Full article

Hedge algebra is a powerful and flexible tool for handling linguistic information, enabling precise quantitative computations and enhancing the effectiveness of multi-criteria decision-making (MCDM). This study proposes a novel integrated fuzzy MCDM approach that combines an enhanced fuzzy analytic hierarchy process (EFAHP) with a 4-tuple hedge algebra semantics model to assess digital transformation in retail enterprises. In this approach, the EFAHP method is integrated with hedge algebra to determine the priorities of pillars and criteria while providing a rigorous mathematical mechanism to transform ambiguous linguistic evaluations into numerical values. This transformation leverages the semantic structure of linguistic variable domains and incorporates fuzziness measures for both atomic words and intensity-modifying words (hedges). Furthermore, a new consistency index formula is introduced to evaluate the reliability of the EFAHP results, with validation being limited to the case study dataset. The 4-tuple hedge algebra semantic model is then employed to assess and rank the digital transformation levels of retail enterprises in Vietnam. Finally, a sensitivity analysis verifies the robustness of the proposed approach by illustrating how variations in pillar and criterion weights influence enterprise rankings. Full article

This study examines the evolving use of social software to combat online disinformation by investigating Weibo users’ attitudes toward IP location disclosure as a measure of transparency and trustworthiness. We analyzed 49,579 posts (April 2022 to May 2023) from Weibo users about IP location disclosure, categorized the topics using LDA topic modeling within the frameworks of communication privacy management, the networked public sphere, and digital democracy, and conducted sentiment analysis. We constructed separate semantic networks for positive and negative terms to examine co-occurrence patterns. The results show that Weibo users are generally negative about this policy, as IP location may reveal personally identifiable information about individuals involved in discussions of online social/political events. Mandatory transparency, while intended to enhance accountability, functions as a mandatory visibility regime that reshapes privacy boundaries and undermines inclusive deliberation. The findings contribute to the exploration of the impact of government-mandatory information privacy disclosure policies on the implementation of platform functionality, as well as changes in user sentiment, information behavior, and components of social media discourse. Full article

University students face rising mental health pressures, making restorative environmental perception (REP) in campus forests critical for psychological recovery. While environmental factors are recognized contributors, Socio-Ecological Systems (SES) theory emphasizes that environmental and social processes are interdependent. Within this context, informal social interaction (ISI)—low-effort encounters such as greetings or small talk—represent a key social dimension that may complement environmental restoration by fostering comfort and embedded affordances. However, most studies examine these factors separately, often using coarse measures that overlook heterogeneity in restorative mechanisms. This study investigates how environmental-exposure and social–environmental context dimensions jointly shape REP in campus forests, focusing on distributional patterns beyond average effects. Using a Public Participation Geographic Information Systems (PPGIS) approach, 30 students photographed 1294 tree-dominant scenes on a forest-rich campus. Environmental features were quantified via semantic segmentation, and ISI was rated alongside REP. Quantile regression estimated effects across the REP distribution. Three distributional patterns emerged. First, blue exposure and ISI acted as reliable resources, consistently enhancing REP with distinct profiles. Second, green exposure functioned as a threshold-dependent resource, with mid-quantile attenuation but amplified contributions in highly restorative scenes. Third, anthropogenic and demographic factors created conditional barriers with distribution-specific effects. Findings demonstrate that campus forest restoration operates through differentiated socio-ecological mechanisms rather than uniform pathways, informing strategies for equitable, restoration-optimized management. More broadly, the distributional framework offers transferable insights for urban forests as socio-ecological infrastructures supporting both human well-being and ecological resilience. Full article

In recent years, food nutrition estimation has received growing attention due to its critical role in dietary analysis and public health. Traditional nutrition assessment methods often rely on manual measurements and expert knowledge, which are time-consuming and not easily scalable. With the advancement of computer vision, RGB-based methods have been proposed, and more recently, RGB-D-based approaches have further improved performance by incorporating depth information to capture spatial cues. While these methods have shown promising results, they still face challenges in complex food scenes, such as limited ability to distinguish visually similar items with different ingredients and insufficient modeling of spatial or semantic relationships. To solve these issues, we propose an Ingredient-Guided Semantic Modeling Network (IGSMNet) for food nutrition estimation. The method introduces an ingredient-guided module that encodes ingredient information using a pre-trained language model and aligns it with visual features via cross-modal attention. At the same time, an internal semantic modeling component is designed to enhance structural understanding through dynamic positional encoding and localized attention, allowing for fine-grained relational reasoning. On the Nutrition5k dataset, our method achieves PMAE values of 12.2% for Calories, 9.4% for Mass, 19.1% for Fat, 18.3% for Carb, and 16.0% for Protein. These results demonstrate that our IGSMNet consistently outperforms existing baselines, validating its effectiveness. Full article