Search Results (1,325)

Low-resolution face recognition poses significant challenges in real-world applications, particularly for deployment on resource-constrained mobile terminals where both accuracy and efficiency are critical. While knowledge distillation has emerged as a promising solution, the existing methods suffer from performance degradation due to negative knowledge transfer when significant capacity gaps exist between teacher and student networks. To address the issue, we propose a novel Hierarchical Neuron Energy Feature Decoupled Distillation (HNEFDD) framework. Firstly, to mitigate the semantic feature gap between teacher and student networks, we integrate a neuron energy feature fusion transfer module into the student network. This module effectively aligns the features of teacher and student networks, alleviating negative knowledge transfer. At the same time, it can adaptively assign higher weights to discriminative facial regions, enabling the network to enhance locally robust structural information and extract semantically richer features in low-resolution face images. Secondly, to achieve efficient and accurate knowledge transfer, we employ a logit-standardized decoupled distillation mechanism that separates the distillation loss from the intermediate-layer auxiliary branches of the teacher and student networks into target-class and non-target-class components. This strategy enables the student network to more accurately learn both the key semantic representations of target-class features and the discriminative boundaries among non-target-class features, thereby achieving efficient and accurate knowledge transfer. Extensive experiments conducted on low-resolution benchmark datasets demonstrate the superiority of the proposed HNEFDD method, which outperforms the current state of the art (SOTA). Full article

Introduction: This study explores obstetric providers’ perceptions of intimate partner violence (IPV) screening in the Chinese clinical setting, guided by the social–ecological model. Methods: Purposive sampling was employed to recruit eight obstetricians and 12 obstetric nurses from a tertiary general hospital in Suzhou, Jiangsu Province, China. Individual semi-structured interviews were conducted between December 2024 and September 2025. Results: Guided by the social–ecological model, this study examines factors influencing healthcare professionals’ engagement in intimate partner violence screening across four levels. A multilevel framework was distilled; four key themes were identified, including the intrapersonal level, interpersonal level, institutional level and community and policy level. (1) The intrapersonal level—healthcare professionals’ competence (knowledge base, communication skills, role identity, screening attitudes, and diagnostic ability); (2) the interpersonal level—interdisciplinary collaboration (internal referrals and external support); (3) the institutional level—hospital screening efficacy (screening resources and management systems); and (4) the community and policy level—social environment (policy support, educational resources, and cultural context). Conclusions: Multilevel factors impact IPV screening in China. These results indicate that a multidimensional model of intervention should be designed to promote IPV screening in China, including integrating IPV curricula into mandatory medical student training, enhancing interdisciplinary collaboration, conducting public education on IPV, implementing socio-cultural adjustments, challenging authoritative attitudes, and providing comprehensive social support. Full article

In strong-flavor Baijiu brewing, surface spraying of recycled distillation tail liquor (TL) leads to volatile aroma loss, uneven substrate distribution and localized fermentation inhibition. A patented telescopic subsurface injector was adopted to deliver 0–30 kg TL per pit at a 50 cm depth on fermentation day 30, the critical transition point between acidogenesis and esterification. At the highest dosage (30 kg), total esters rose 14.8% (from 4.45 to 5.11 g/L), with ethyl hexanoate up 52.2% (to 1.72 g/L) and ethyl lactate up 69.8% (to 3.82 g/L). Fermentation temperature curves stayed unchanged, and grain-derived ethanol yield remained near 38.2% in all groups after correcting for the ethanol already present in the added TL. Subsurface injection of tail liquor-supplying ethanol and organic acid precursors-at the acidogenesis-to-esterification transition enhances ester synthesis without disrupting fermentation or reducing distillate yield. Acid profiles were also stable, suggesting that the added substrates were channeled into ester synthesis rather than acid accumulation. Targeted subsurface TL injection at this metabolic transition thus represents an industrially feasible strategy to boost ester biosynthesis, offering a recyclable TL valorization approach and verifying substrate-limited esterification in solid-state Baijiu fermentation. Full article

Low temperature during germination of late-seeded rapeseed disrupts multiple physiological and biochemical processes and thus limits growth and yield. Accordingly, methods to improve cold tolerance in late-sown rapeseed are needed. In this study, Zhongshuang 11 seeds were primed for 10 h with different concentrations of erucic acid (EA) or glucosinolates (GSLs). After drying, seeds were germinated at low temperature (15 °C/10 °C, 16 h/8 h light/dark) for 14 days. Compared with the control (distilled water priming), the optimal treatments—500 mg/L EA and 300 mg/L GSLs—increased germination rates by 2.9% and 15.6%, respectively, and raised total seedling biomass by 14–24%. Physiological assays on day 14 showed that EA priming increased peroxidase (POD) activity by 28.3%, while GSL priming enhanced superoxide dismutase (SOD) and POD activities by 12.6% and 36.2%, respectively. EA seed priming increased auxin (IAA), brassinolide (BR), cytokinin (CTK), and gibberellin (GA) contents in underground tissues by 37.2%, 18.7%, 53.9%, and 46.7%, respectively, while GSL priming raised IAA, BR, and GA levels in aerial tissues by 74.0%, 59.0%, and 26.6%. Moreover, EA seed priming significantly increased the activities of long-chain acyl-CoA synthetase (LACS) and carnitine acyltransferase (CPT) in rapeseed seedlings, whereas GSL priming elevated glutathione S-transferase (GST) and thioredoxin reductase (TrxR) activities. Field experiments confirmed that EA and GSL priming enhanced seedling biomass accumulation, producing 31.4% and 23.8% increases in total dry weight, respectively, and increased silique number per plant by 15.6% and 17.3%, ultimately raising grain yield by 12.9% and 20.0%. These results indicate that EA or GSL seed priming can improve cold tolerance and yield of late-seeded rapeseed, although further multi-environment and mechanistic studies are required. Full article

Membrane distillation (MD) technology has become an effective solution to freshwater scarcity due to its low energy consumption, high separation efficiency, and ability to handle highly concentrated saline wastewater. Nevertheless, issues such as membrane wetting, membrane fouling, and low membrane flux severely limit its large-scale application. Composite membranes prepared using metal–organic framework (MOF) materials as fillers have become a research hotspot due to their advantages, such as permeable microporous channels, customizable pore structures, and modifiable active sites. These properties enable them to effectively reduce temperature polarization and concentration polarization phenomena. This article describes the characteristics of MOF materials and their current applications in the field of MD, with a comparative analysis of the applicability of MOF polycrystalline membranes and MOF composite membranes in MD, and discusses the working principle of MOFs in enhancing the performance of MD. Finally, the problems and challenges associated with the use of MOFs in MD applications are analyzed. This study aims to provide theoretical guidance for the application of MOF materials in the field of MD seawater desalination. Full article

Industrial surface defect detection faces challenges of complex textures, diverse defect morphologies, and scarce labeled data, especially for non-woven fabrics. This paper proposes a dual-domain reverse distillation algorithm for unsupervised defect detection (DDRD). The algorithm integrates frequency-domain wavelet enhancement and spatial-domain self-attention to enhance defect features synergistically. A Wavelet High-frequency Deformable Enhancement Module amplifies fine-grained defect details, while a Convolutional Self-Attention Spatial Global Enhancement Module captures long-range spatial dependencies. The enhanced dual-domain features are embedded into a reverse distillation framework for end-to-end training, achieving precise defect localization via feature reconstruction errors. Extensive experiments are conducted on the self-built large-scale non-woven fabric WFB dataset and the public MVTec AD benchmark dataset. The results show that DDRD achieves 98.0% pixel-level AUROC and 93.0% image-level AUROC on the WFB dataset, outperforming the state-of-the-art RD4AD method by 0.5% and 1.5%, respectively. On the MVTec AD dataset, it attains an average of 97.5% pixel-level AUROC and 99.7% image-level AUROC, with perfect 100% image-level detection accuracy on multiple categories. These results validate the efficacy and robustness of the dual-domain enhancement strategy for industrial defect detection tasks. Full article

Water scarcity is a major constraint to sustainable horticultural production, particularly in arid and semi-arid regions. This study evaluated the effectiveness of two natural foliar biostimulants, chitosan and Spirulina extract, in improving water productivity, growth performance, biomass production, and drought tolerance of passion fruit (Passiflora edulis Sims) seedlings under deficit irrigation. A factorial experiment was conducted using three irrigation regimes (100%, 85%, and 70% of field capacity, FC) combined with three foliar spray treatments: distilled water (control), chitosan (1 g L⁻¹), and Spirulina extract (1 g L⁻¹). Deficit irrigation significantly reduced chlorophyll content, vegetative growth, biomass production, and water productivity, with the greatest reductions observed at 70% FC. Foliar application of both biostimulants effectively alleviated these adverse effects by enhancing plant growth, biomass production, chlorophyll retention, and water productivity under water-limited conditions. Chitosan consistently outperformed Spirulina, particularly under moderate water deficit, by reducing irrigation water requirements while maintaining seedling growth and physiological performance comparable to those under full irrigation. Multivariate analyses, including Pearson correlation, principal component analysis, and response surface methodology, revealed strong positive relationships among biomass production, chlorophyll content, vegetative growth traits, and water productivity. Among all treatments, foliar application of chitosan at 1 g L⁻¹ combined with 85% FC was identified as the optimum strategy, providing substantial irrigation water savings while maintaining vigorous seedling growth, high biomass production, and superior water productivity. These findings demonstrate that integrating moderate deficit irrigation with chitosan application is an effective and sustainable strategy for improving passion fruit nursery production under conditions of limited water availability. Full article

Existing research on e-commerce product reviews has primarily focused on analysing consumers’ opinions, emotions, sentiments and associated star ratings. Whilst these approaches provide insights into consumers’ perceptions of products, they offer limited understanding of how products are used in real-world contexts. Therefore, they do little to enhance the e-commerce experience by helping consumers make more informed purchasing decisions based on products’ intended uses without requiring them to read numerous reviews during the decision-making process. To address this problem, this paper investigates the feasibility of automatically identifying and classifying product usage activities from e-commerce reviews. A methodology combining natural language processing, manual activity-level annotation and deep learning-based text classification was developed and evaluated. An initial dataset of 60,000 Amazon product reviews was manually labelled according to six activity classes: run, walk, hike, swim, climb and unknown. Following quality inspection and data cleaning, a final dataset of 50,843 reviews was used for model training and evaluation. Multiple classification approaches were assessed, including CNN, LSTM, hybrid LSTM-CNN architectures and transformer-based models (DistilBERT and DistilBERT-CNN). Experimental evaluation was conducted using multiple random seeds to ensure robustness and reproducibility. The results indicate that activity classification from e-commerce reviews is a challenging task due to ambiguity and overlapping usage descriptions, with all evaluated models achieving comparable performance on the full dataset. Among the evaluated models, the hybrid LSTM-CNN-GloVe architecture achieved the highest performance on the keyword-filtered dataset, whilst the DistilBERT-CNN model also demonstrated strong results. The findings demonstrate the feasibility of extracting activity-oriented information from product reviews and highlight activity classification as a distinct and under-explored natural language processing task that complements traditional sentiment analysis. The proposed methodology provides a foundation for improving product discovery and supporting usage-oriented search and recommendation systems in e-commerce environments. Full article

To enhance unmanned ground vehicle (UGV) intelligence in smart cities, disaster rescue, and infrastructure inspection, this paper investigates the collaborative optimization of multimodal fusion end-to-end architectures. Through dynamic alignment of heterogeneous features, a multi-head distillation attention mechanism, and parallel decision-planning, a high-precision, low-latency closed-loop autonomous navigation framework is constructed. A Multi-Head Distillation Attention-based Trajectory Prediction (MDA-TP) method is proposed, combining a BEVFormer-based multimodal fusion perception model with a two-stage progressive knowledge distillation framework. On NuScenes, the method achieves an ADE of 0.88 m, an FDE of 1.32 m (4.34% and 10.81% reductions), and a collision rate of 15.2%, with 42.6 M parameters and 46 ms latency. Ablation shows removing attention distillation increases FDE by 13.6%. For system validation, a multi-sensor UGV platform is built. Through NuScenes online testing and real-world closed-loop validation, the Euclidean deviation remains within 0.5 m. Compared with traditional distillation, speed prediction MSE is reduced by 51.5%, wheel angle RMSE by 58.4%, and route completion improves from 60.99% to 97.26%. The results provide practical support for autonomous driving in smart cities, disaster rescue, and infrastructure inspection. Full article

Urban infrastructure systems increasingly depend on textual reports generated by citizens, inspection teams, maintenance units, emergency platforms, and smart city services. Accurate identification of critical risks in these reports is essential for enhancing urban resilience and enabling timely decision-making. Nevertheless, urban infrastructure risk classification is challenging due to the brevity, noise, domain specificity, and context dependence of these reports. This study introduces RAFE-XAI, a retrieval-augmented feature engineering and explainable natural language processing framework for urban infrastructure risk classification. The term retrieval-augmented is used here in a classification-oriented sense: retrieved reports are used to construct additional features and evidence, not to generate output text as in Retrieval-Augmented Generation systems. The proposed framework incorporates semantic sentence embeddings, retrieval-based evidence, neighborhood-derived label distributions, domain-specific risk indicators, infrastructure asset cues, location indicators, and evidence-based explainability. The framework does not construct an explicit graph, adjacency matrix, graph neural network, or message-passing mechanism. Instead, retrieval is used to derive neighbor label-distribution features, which are combined with semantic embeddings and interpretable keyword, asset, and location indicators. To assess the effectiveness of this approach, UIR-Text, a semi-synthetic urban infrastructure risk dataset with scenario-level group splitting to mitigate data leakage, was constructed. Experimental results on UIR-Text show that fine-tuned DistilBERT achieves the strongest predictive performance, with Macro-F1 scores of 0.8278 for category classification, 0.9120 for binary critical-risk detection, and 0.3379 for four-level severity classification. Among the explainable feature-engineering models, RAFE-XAI with Random Forest achieves the strongest category classification performance, with Accuracy 0.8400, Macro-F1 0.8043, Weighted-F1 0.8444, and MCC 0.8062. These results suggest that fine-tuned transformers provide the highest predictive performance on this benchmark, while RAFE-XAI offers a transparent retrieval-augmented alternative that exposes retrieved evidence, neighbor label distributions, and domain cues. Four-level severity classification remains challenging, even with fine-tuned DistilBERT, indicating the need for richer impact-aware variables. Full article

Mixed food waste represents a realistic yet highly heterogeneous feedstock for bioethanol production; however, most existing studies focus on homogeneous waste streams or isolated process stages such as hydrolysis, fermentation, or distillation. This fragmented approach limits a comprehensive understanding of overall system performance, particularly when feedstock variability, process selection, and energy demand are considered simultaneously. In this study, a comparative assessment of bioethanol feedstocks and pretreatment strategies is first conducted to identify mixed food waste as a representative substrate and to determine suitable conversion pathways. Building on this, an integrated process simulation and optimization framework is developed using Aspen HYSYS to model the complete bioethanol production chain, including pretreatment, enzymatic liquefaction, saccharification, fermentation, and downstream purification. The novelty of this work lies in the system-level optimization of the entire process rather than individual unit operations, combined with the implementation of heat integration and process recycling strategies to enhance overall efficiency. Optimal operating conditions were identified at 72 °C for liquefaction, 57 °C for saccharification, and 32 °C for fermentation. In the separation section, a condenser temperature of 79 °C, a separator temperature of 20 °C, and a 25-stage final distillation column enabled efficient ethanol recovery with a purity of 98.9%. Heat integration reduced total utility demand by 87.25%, while recycling streams improved resource utilization and minimized process waste. Overall, the study demonstrates that combining feedstock-level assessment with system-level optimization provides a robust, efficient, and scalable pathway for sustainable bioethanol production from heterogeneous waste resources. Full article

Optical remote sensing object detection faces challenges such as large variations in scale, slender and direction-sensitive targets, complex backgrounds, and limited deployment resources. This paper proposes a lightweight geometrically decoupled student network with a dynamic multi-teacher distillation framework. Based on YOLO11n, the student detector keeps the original classification branch while redesigning the regression branches at different scales. Lightweight regression towers are used for the shallow and deep branches, whereas a geometrically decoupled regression tower is introduced only at the intermediate branch to enhance localization for slender and orientation-sensitive objects with limited extra cost. A geometry-adaptive box loss is further employed to stabilize localization training. For knowledge transfer, three specialized teachers are constructed for semantic classification, geometric regression, and structural topology supervision. A branch-decoupled adaptive weighting strategy dynamically integrates their complementary knowledge for classification and regression distillation. Experiments on DIOR show that the proposed model reduces parameters by 6.8% and GFLOPs by 13.9%, while improving mAP50 by 0.23 percentage points over YOLO11n. Validation on NWPU VHR-10 and deployment tests using PT, ONNX, and TensorRT further demonstrate improved accuracy–efficiency trade-offs and practical inference acceleration. Full article

Detecting AI-generated text has become a critical task as artificial intelligence systems are increasingly used in content creation. Current detection methods often suffer from limited accuracy and weak multilingual performance. This problem is especially challenging in Turkish, Arabic, and English due to their distinct linguistic structures, including agglutinative morphology in Turkish, root-based morphology in Arabic, and semantic ambiguity in English. To address these challenges, this study proposes a hybrid architecture that combines a Transformer-based DistilBERT model with a Graph Convolutional Network (GCN). While DistilBERT captures rich contextual and semantic information, GCN enhances detection by modeling structural relationships within text data. The proposed model is evaluated against other well-known approaches. Experimental results show that the hybrid DistilBERTGCN framework achieves high detection accuracy, reaching 99% for English and 98% for Turkish and Arabic. In addition, this study introduces new multilingual datasets, contributing to the advancement of the literature research. Full article

Generative data augmentation based on diffusion models has emerged as a promising approach for few-shot image classification. Existing methods, such as DA-Fusion, typically follow a “generate-once, use-directly” paradigm, which often suffers from uncontrollable generation quality, unstable semantic consistency, insufficient global diversity, and high sample redundancy. To address these limitations, we propose a two-stage Propose-and-Select framework for controllable data augmentation. This framework curates high-quality synthetic data offline, ensuring that no additional training overhead is introduced to downstream models. For selector optimization, our method eliminates the need for additional human annotations by leveraging the zero-shot prior knowledge of a vision–language model (CLIP) to construct relative-quality pseudo-labels. Furthermore, we develop an adaptive-temperature listwise ranking distillation objective to transfer quality-aware supervision effectively. We also introduce a multi-objective consistency regularization strategy to stabilize training and improve convergence. Under a strictly controlled augmentation budget, where all methods are provided with the same number of synthetic samples, the proposed approach consistently outperforms existing diffusion-based augmentation baselines across both few-shot classification benchmarks, achieving an accuracy of 79.58% on PASCAL VOC and 80.74% on the fine-grained Oxford 102 Flowers dataset. These results demonstrate the effectiveness of the proposed generation-selection paradigm in improving the quality, diversity, and semantic relevance of synthetic samples, thereby enhancing downstream few-shot classification performance. Full article

Background: Parkinson’s disease (PD) is a progressive neurodegenerative disorder characterized by the selective loss of dopaminergic neurons in the substantia nigra. Oxidative stress, neuroinflammation, and α-synuclein aggregation are central pathological features of PD. Zanthoxylum piperitum DC, commonly known as Korean pepper or chopi, is a traditional dietary spice in Eastern Asia and has been reported to possess antioxidant and anti-inflammatory properties. This study investigated the neuroprotective and motor function–enhancing effects of distilled extract of Z. piperitum (deZP) in 1-Methyl-4-phenylpyridinium (MPP⁺)-treated Caenorhabditis elegans and 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP)-induced mouse models of PD. Methods: In the C. elegans model, dopaminergic neurotoxicity was induced by MPP⁺, and deZP was tested at 0.25, 0.5, and 1% (v/v) to evaluate neuronal preservation through GFP-labeled dopaminergic neurons and α-synuclein expression. Concurrently, in the MPTP-induced mouse model, deZP was administered intranasally at a fixed dose of 20 μL/mouse, equivalent to 5 mg/mouse. Motor function was assessed using the rota-rod test, pole test, and grip strength test, while dopaminergic neuronal survival was evaluated by tyrosine hydroxylase (TH) immunostaining. Results: In MPP⁺-treated C. elegans, deZP significantly restored green fluorescent protein (GFP) fluorescence in dopaminergic neurons and reduced α-synuclein expression, with the most pronounced effects observed at 1% (v/v). In the MPTP-induced mouse model, deZP at this fixed intranasal dose significantly improved motor performance and preserved TH-positive neurons in the substantia nigra. Conclusions: These findings suggest that deZP may represent a promising preclinical candidate for further investigation in PD-related neurodegeneration. Full article