Search Results (1,113)

Thermal monocular depth estimation can provide more robust depth predictions than RGB-based methods under nighttime and adverse weather conditions. However, when trained with projected LiDAR supervision, depth models often retain structural errors in sky regions, long-range areas, and object boundaries because LiDAR measurements are sparse or missing in such regions. To address this limitation, we propose a thermal monocular depth estimation framework that incorporates propagation-based refinement. To make this refinement applicable across different base models, we further design a multi-scale feature adapter that converts heterogeneous multi-scale features with different spatial resolutions and channel dimensions into a unified representation. As a result, the same refinement architecture can be used across different base models without model-specific refiner redesign. On the multispectral stereo (MS2) dataset, the proposed method improves both BTS (big-to-small) and NeWCRFs (neural window fully connected CRFs), reducing the meter-based error metrics SqRel from 0.380 to 0.369 and RMSE from 3.163 to 3.126 for BTS, and reducing SqRel from 0.331 to 0.328 and RMSE from 2.937 to 2.924 for NeWCRFs. Qualitative results further show that the proposed method alleviates mixed-depth artifacts and abnormal depth patterns in regions lacking reliable depth supervision. Full article

Land use change detection (LUCD) serves as a crucial technical cornerstone for natural resource management and ecological environment monitoring, playing an indispensable role in advancing the modernization of national governance capacities. Nonetheless, severe interference from radiometric variations on feature representation readily induces spurious changes and thus a high false alarm rate. Additionally, the challenge of balancing discriminative feature extraction and fine-grained contextual modeling leads to fragmented change regions and missed detection. To address these issues and eliminate the reliance on annotated samples, a novel framework is proposed for unsupervised LUCD, integrating superpixel-based deep feature analysis with a dense conditional random field (CRF). Firstly, relative radiometric correction and band-wise maximum stacking fusion are performed on the bi-temporal images. A simple non-iterative clustering (SNIC) algorithm is adopted to generate homogeneous superpixels with cross-temporal consistency. Then, a deep feature coupling mining mechanism is introduced to implement spatial–spectral feature extraction and in-depth parsing of invariant semantic information. Meanwhile, the difference confidence map based on dual features is constructed using superpixel-level discriminant vectors to enhance the separability. Finally, leveraging homogeneous units with spatial correspondence, a task-specific redesign of a global optimization model is established to achieve the precise extraction of change regions, which incorporates difference confidence, spatial adjacency relationship, and cross-temporal feature similarity into the dense CRF. The experimental results demonstrate that the proposed method achieves an average overall accuracy of over 90% across all datasets with excellent comprehensive performance, striking a well-balanced trade-off in practical applicability. It can effectively suppress salt-and-pepper noise, significantly improve the recall rate of change regions (maintaining at approximately 90%), and exhibit favorable superiority and robustness in complex land cover scenarios. Full article

This study investigates the decomposition kinetics and microplastic residue formation of the polymer-coated controlled-release fertilizers (CRFs) LN40 and Eco-LN40 under simulated photodegradation conditions. Eco-LN40, containing TiO₂ as a photocatalyst, achieved complete decomposition (100 ± 2%) after 60 days of xenon-arc irradiation (p < 0.05), whereas LN40 achieved only 14–31% decomposition. Analytical characterization using TED-GC/MS, FTIR, and Raman spectroscopy confirmed that polyethylene (PE) signals completely disappeared in Eco-LN40 but persisted in LN40, indicating that microplastics did not form and that there was total oxidation into CO₂ and H₂O. SEM–EDS revealed Ti enrichment and surface fragmentation consistent with photoinduced radical oxidation. This study provides qualitative and mechanistic evidence that TiO-catalyzed photodegradation can eliminate polymer residues, mitigate the risk of microplastic contamination in agricultural soils, and support carbon-neutral fertilizer technologies. Full article

To deal with the high standards, strong restrictions, and high repeatability that are inside State Grid small-scale infrastructure projects, this research puts forward a norm-driven generative design method, which conquers the low efficiency, compliance dangers, and semantic breakage that are usual in manual modeling. Taking standards such as Q/GDW 11382.3-2015 as the knowledge origin, we construct an ALBERT-BiLSTM-CRF semantic parsing model and change natural-language clauses into executable design restrictions via normative text pre-processing, BIO sequence marking, and rule triplet mapping. Therefore, model training and assessment produce Accuracy, Precision, Recall, and F1 of 98.05%, 95.49%, 95.88%, and 95.59% separately, with 100% precision for logical comparison and conjunction labels; thus, this provides a steady semantic base for the rule base. At the component level, a three-part coding plan and unit module collection are built based on OmniClass and GB/T 51269, which makes semantic consistency and traceability between components and space functions possible. At the system level, a continuous work process is carried out through the Revit API, which covers scheme making, automatic arrangement, and deliverable output. Hence, validation on a real case in a digital operation center for the power system shows that the design time for the third-floor administrative office area was cut from about 20 h to around 4 h, and the first-time solution met all code restrictions, which improves efficiency and compliance in a significant way. The results point out that norm-driven generative design can supply deployable automation and high-quality outputs for small-scale power infrastructure, which provides a sustainable database for digital twins and smart O&M. Full article

Lightweight intrusion detection systems (IDSs) are increasingly integrated into applied data science workflows for cybersecurity and process monitoring, where limited computational resources and asymmetric error costs constrain model design. This paper presents a comparative study of two lightweight deep learning IDS architectures: ForNet, a convolutional model optimized for feature-centric detection, and SigNet, a gated recurrent model designed for sequence-oriented modeling of ordered flow-feature representations. Both models are trained with Cost-Robust Focal Loss (CRF-Loss), a cost-aware objective that penalizes false positives and false negatives according to deployment-specific risk preferences. We evaluate the models on the UNSW-NB15 and CIC-IDS2017 benchmarks using six standard metrics (accuracy, precision, recall, F1-score, Matthews correlation coefficient (MCC), and the area under the receiver operating characteristic curve (AUROC)), complemented by an analysis of false-positive behavior. On CIC-IDS2017, ForNet achieves precision up to 0.95 and MCC up to 0.93 with AUROC above 0.94, while SigNet shows a stronger recall-oriented profile on UNSW-NB15. In an ablation study, replacing Binary Cross-Entropy with CRF-Loss reduces the false-positive rate by approximately 15–20% and improves robustness-oriented metrics such as MCC by up to 12% on CIC-IDS2017. Rather than claiming universal state-of-the-art performance, the study focuses on performance–risk trade-offs under realistic operational constraints. The results highlight how architectural bias and cost-aware optimisation jointly shape IDS behaviour and offer benchmark-based guidance for interpreting performance–risk trade-offs in lightweight intrusion detection. Full article

Kazakh sequence labeling is constrained by limited annotated resources, while its agglutinative morphology and productive suffixation increase data sparsity and exacerbate label inconsistency in part-of-speech (POS) tagging and named entity recognition (NER). This paper proposes an LLM-assisted weak supervision framework in which a large language model generates synthetic token-level annotations that are subsequently filtered using confidence-based criteria and combined with a smaller manually verified subset to train Transformer-based sequence taggers with Conditional Random Field (CRF) decoding. The pipeline unifies corpus construction, weak-label generation, quality filtering, word-to-subword alignment, and CRF-refined structured prediction into a reproducible workflow. Experimental results show that contextual encoders and structured decoding provide strong performance for Kazakh POS and NER, while the proposed training design enables efficient convergence with diminishing returns beyond moderate epoch budgets. Error-slice analysis indicates that residual errors are concentrated in rare tokens, morphologically complex long words, longer sentences, and the ORG entity class. Overall, the findings support the use of LLM-assisted weak supervision as a scalable strategy for low-resource Kazakh sequence labeling when synthetic labels are controlled through filtering and refined by structured decoding. Full article

Quantitative microstructure evaluation of sedimentary rock thin sections is essential for revealing reservoir flow mechanisms and assessing reservoir quality. However, traditional manual identification is inefficient and prone to subjectivity. Although current deep learning approaches have improved efficiency, most remain confined to single tasks and lack a pathway to translate image recognition into quantifiable geological parameters. Moreover, these methods struggle with cross-scale feature extraction and accurate grain boundary localization in complex textures. To overcome these limitations, this study proposes a three-stage automated analysis framework integrating intelligent lithology identification, sandstone grain segmentation, and quantitative analysis of fabric parameters. To address scale discrepancies in lithology discrimination, Rock-PLionNet integrates a Partial-to-Whole Context Fusion (PWC-Fusion) module and the Lion optimizer, which mitigates cross-scale feature inconsistencies and enables accurate screening of target sandstone samples. Subsequently, to correct boundary deviations caused by low contrast and grain adhesion, the PetroSAM-CRF strategy integrates polarization-aware enhancement with dense conditional random field (DenseCRF)-based probabilistic refinement to extract precise grain contours. Based on these outputs, the framework automatically calculates key fabric parameters, including grain size and roundness. Experiments on 3290 original multi-source thin-section images show that Rock-PLionNet achieves a classification accuracy of 96.57% on the test set. Furthermore, PetroSAM-CRF reduces segmentation bias observed in general-purpose models under complex texture conditions, enabling accurate parameter estimation with a roundness error of 2.83%. Overall, this study presents an intelligent workflow linking microscopic image recognition with quantitative analysis of geological fabric parameters, providing a practical pathway for digital petrographic evaluation in hydrocarbon exploration. Full article

The breeding blanket is a key component of tokamaks, primarily responsible for extracting heat from fusion reactions and for tritium breeding, which is essential to ensure a fusion reactor’s fuel self-sufficiency. Recent technological advancements have led to the development of Dual-Cooled Lead–Lithium (DCLL) breeding blankets, which employ a liquid metal (specifically a Lead–Lithium eutectic alloy) as a heat transfer medium and tritium breeder, while helium gas is used to cool the structural components of the reactor. The interaction between the moving electrically conducting fluid and the strong magnetic field in the tokamak environment leads to magnetohydrodynamic (MHD) effects. The latter are characterized by the induction of eddy currents within the fluid and resulting Lorentz forces generated by their interaction with the magnetic field, which cause additional pressure losses and reduce heat transfer efficiency. This work investigates the pressure drop experienced by a Lead–Lithium flow within a rectangular section conduit under the action of an external, uniform magnetic field of different intensities. An analytical model was developed to estimate the total MHD-induced pressure losses along the channel for different values of the external magnetic field intensity and then benchmarked against relative computational fluid dynamics (CFD) simulations carried out using COMSOL Multiphysics. This comparison allowed the validation of the analytical predictions as well as a better understanding of the influence of the applied magnetic field intensity on the overall pressure drop. Therefore, the aim of the analytical model is to provide analytical tools for reasonably accurate estimations of MHD pressure losses suitable for future preliminary design purposes. Full article

Named Entity Recognition (NER) plays a crucial role in Aspect-Based Sentiment Identification (ABSI), enabling the extraction of domain-specific aspects and their associated sentiment expressions from unstructured textual reviews. In complex domains such as movie reviews, sentiment is frequently conveyed through references to named entities (e.g., actors, directors, or movie titles) and other contextual cues. However, many existing ABSI approaches treat NER as a separate preprocessing step, limiting the effective modeling of entity–aspect–opinion relationships. Integrating NER directly into the ABSI framework, allows entity-specific opinions to be more accurately identified, overlapping aspects to be disambiguated, and contextual sentiment expressions to be captured more effectively. To address these challenges, this study proposes an integrated NER-based aspect identification model built on feature-enhanced LSTM and BiLSTM architectures. Linguistic features, including Parts-of-Speech (POS) tags and chunking information, are incorporated to enrich contextual representations, while a Conditional Random Field (CRF) decoding layer models inter-label dependencies for coherent sequence-level predictions of named entities, aspects, and associated opinion expressions. Compared with large transformer-based models, the proposed BiLSTM-CRF architecture offers lower computational complexity, fewer parameters, and allows explicit integration and analysis of linguistic features that are often implicitly encoded in transformer attention mechanisms. The model is evaluated through multiple experimental variants across three domains. Four configurations are applied to movie-review data to jointly extract person names, movie titles, and aspect-opinion pairs, while six configurations assess cross-domain robustness on restaurant and laptop review datasets. Results show that the BiLSTM-CRF model augmented with POS features consistently outperforms baseline configurations in the movie domain and remains competitive across domains, achieving an F1-score of 0.89. These findings demonstrate that explicit linguistic feature integration within a CRF-based sequence modeling can provide an effective and computationally efficient alternative to large-scale transformer fine-tuning for structured, entity-linked ABSI tasks. Full article

To enhance flight safety and operational efficiency, extracting key actions, flight parameters, and status information from civil aviation flight crew instructions generated during pre-flight and in-flight procedures is crucial. However, such texts are highly condensed and involve complex multi-role interactions, easily leading to entity boundary drift and category misclassification. To address this, this paper proposes a joint key information extraction framework based on a lightweight pre-trained language model (TinyBERT) and a Role-Aware Fusion mechanism, abbreviated as TR-BiGRU-CRF. This framework introduces the Role-Aware Fusion mechanism to resolve semantic ambiguity caused by multi-party interactions, utilizes TinyBERT for semantic representation that balances accuracy and computational efficiency, and employs BiGRU-CRF for robust sequence feature modeling and decoding. Experiments on a flight crew instruction dataset show that the proposed method achieves 92.2% precision, 91.8% recall, a 92.0% F1 score, and an overall prediction accuracy of 92.6%. Compared to the BiGRU-CRF baseline, it significantly improves accuracy, precision, and F1 score by 11.4, 13.3, and 13.5 percentage points, respectively. These results prove that the proposed method effectively mitigates boundary drift and category confusion, providing strong support for flight crew instruction understanding and safety decision-making. Full article

Background/Objectives: Cervical facet-mediated pain is a significant underlying factor of persistent neck pain (CNP). Pulsed radiofrequency (PRF) or conventional radiofrequency (CRF) has been introduced as a treatment alternative. However, comparative clinical data remain limited. Methods: This single-center retrospective cohort study analyzed patients with cervical facet-mediated pain who underwent PRF (n = 40) or CRF (n = 44) between January 2023 and December 2024. The success of the procedure was assessed using the Numeric Rating Scale (NRS) and the Neck Disability Index (NDI) before the procedures and at 1, 3, 6, and 12 months following the injections. Patients’ feedback was evaluated using the Global Perceived Effect (GPE) scale. Results: For both groups, a substantial decrease in the mean pain and disability severity was recorded between the initial measurement and the first, third, and sixth months of follow-up, but the outcomes were significant only in the CRF group at the 12th month. The groups did not show a substantial difference in terms of pain relief, disability improvement, medication use, or patient satisfaction at one and three months (p > 0.05), but at six and 12 months, patients treated with CRF showed significantly greater outcomes (p < 0.001). No notable difference in complication rates was found between the PRF (10%) and CRF (16%) groups (p = 0.53). Conclusions: Both pulsed and conventional radiofrequency ablation effectively reduced pain and improved function in the early-midterm follow-up. However, CRF provided more sustained relief and greater patient-reported success at 6 and 12 months, without an increase in complication rates, suggesting that CRF may offer a more durable long-term treatment option. Full article

Human Immunodeficiency Virus Type 1 (HIV-1), the causative agent of the acquired immune deficiency syndrome (AIDS), originated from zoonotic transmissions of simian immunodeficiency viruses (SIVs) infecting African great apes, following complex cross-species transmission events and virus–host co-evolution. These processes were accompanied by multiple viral adaptations, particularly within structural and accessory genes, enabling evasion of host restriction factors and long-term viral persistence. In 1988, an antisense open reading frame (ORF) overlapping the env gene was proposed and subsequently confirmed by the identification of antisense transcripts and the antisense protein (ASP). An “intact” ASP ORF (defined as >150 codons) is predominantly conserved in pandemic HIV-1 group M viruses and shows evidence of positive selection, suggesting a selective advantage. Increasing evidence supports the hypothesis that the asp gene emerged de novo during the evolution of group M and contributed to viral adaptation and global spread in humans. This review combines a narrative review of the literature with original in silico analyses of HIV-1 and SIV sequences retrieved from the Los Alamos National Laboratory database. We systematically reassessed the distribution, length variability and conservation of the ASP ORF across HIV-1 groups (M, N, O, P), subtypes, circulating recombinant forms (CRFs), unique recombinant forms (URFs) and related SIV lineages. Our updated analyses confirmed the strong association between the presence of an “intact” ASP ORF and pandemic HIV-1 group M lineages, while revealing rare but notable antisense ORFs in selected SIVcpz and SIVgor strains. By integrating evolutionary, epidemiological and sequence-based evidence, we aim to clarify the origin and maintenance of the ASP ORF and to contextualize its emergence within the broader framework of overlapping gene evolution, de novo gene birth and the selective pressures shaping viral fitness and pandemic potential. Full article

Multimodal named entity recognition (MNER) leverages both textual and visual information to improve entity recognition, particularly in unstructured scenarios such as social media. While existing approaches predominantly rely on raster images (e.g., JPEG, PNG), scalable vector graphics (SVG) offer unique advantages in resolution independence and structured semantic representation—an underexplored potential in multimodal learning. To fill this gap, we propose MNER-SVG, the first framework that incorporates SVG as a visual modality and enhances it with ChatGPT-generated auxiliary knowledge. Specifically, we introduce a Multimodal Similar Instance Perception Module that retrieves semantically relevant examples and prompts ChatGPT to generate contextual explanations. We further construct a Full-Text Graph and a Multimodal Interaction Graph, which are processed via Graph Attention Networks (GATs) to achieve fine-grained cross-modal alignment and feature fusion. Finally, a Conditional Random Field (CRF) layer is employed for structured decoding. To support evaluation, we present SvgNER, the first MNER dataset annotated with SVG-specific visual content. Extensive experiments demonstrate that MNER-SVG achieves state-of-the-art performance with an F1 score of 82.23%, significantly outperforming both text-only and existing multimodal baselines. This work validates the feasibility and potential of integrating vector graphics and large language model-generated knowledge into multimodal NER, opening a new research direction for structured visual semantics in fine-grained multimodal understanding. Full article

Fault diagnosis in modern power systems increasingly depends on unstructured operation and maintenance (O&M) logs, yet real-world logs are often small in scale and highly imbalanced across fault types, which degrades the generalizability of standard neural models. This paper proposes a graph-augmented diagnostic framework that integrates imbalance-aware knowledge extraction with interpretable reasoning. The framework consists of three stages: (1) domain adaptation of a BERT–BiLSTM–CRF NER model and a BERT–MLP RE model using an imbalance-aware training recipe that combines Low-Rank Adaptation (LoRA), a mixed focal–range loss, and undersampling; (2) construction of a power-system knowledge graph that organizes extracted entities and relations (e.g., fault devices, abnormal phenomena, causes, and handling measures); and (3) a graph-augmented assistant agent that reuses the NER model as a graph-aware retriever within a retrieval-augmented generation (RAG) architecture to support contextualized and interpretable diagnostic reasoning. Experiments on 3921 real-world fault-processing logs show consistent gains: NER reaches 92.0% accuracy and 71.3% Macro-F1 (vs. 80.3% and 63.2%), and RE achieves 88.0% accuracy and 70.1% F1 (vs. 82.1% and 60.4%), while reducing average training time per epoch by about 18%. These results demonstrate an efficient and practical path toward robust log-based fault diagnosis under scarce and imbalanced data. Full article

Objective: Population mobility complicates the prevention and control of HIV. To address these challenges, this study explored the molecular epidemiology of HIV among migrant populations in Hangzhou. Methods: People newly diagnosed with HIV/AIDS from 2020 to 2023 were divided into permanent migrant population (PMP), temporary migrant population (TMP), and non-migrant population (NMP). HIV-1 pol gene sequencing was performed to calculate genetic distance. Sample pairs with genetic distances ≤0.005 were used to construct the molecular transmission network. Results: PMP comprised people living with HIV in Hangzhou, characterized by younger age, higher education, and predominantly homosexual transmission. This population forms multiple large molecular clusters together with NMP. TMP accounted for the highest proportion of females and people infected through heterosexual contact, but the education level was the lowest. NMP had the fewest people living with HIV. The main subtypes identified were CRF01_AE, CRF07_BC, CRF08_BC and CRF55_01B. Drug resistance prevalence did not differ significantly among the populations. The molecular transmission network included 833 cases forming 275 clusters, with an overall sample inclusion rate of 23.04%. PMP, TMP and NMP inclusion rates were 27.10%, 19.03% and 21.4%, respectively. All molecular clusters involved migrant populations. Factors associated with inclusion in the network for migrants included current residence, household registration, STD history, sample source, and stage at diagnosis. Conclusions: Migrant populations play a major role in ongoing HIV transmission. Prevention and control measures should be strengthened according to population-specific characteristics. Molecular transmission networks are useful tools for assisting precise control. Full article