Search Results (674)

Reliable estimation of humanoid robot body configuration from monocular RGB frames is useful for external monitoring in traffic, logistics, and service-robotics environments. However, direct transfer of human pose models is challenged by differences in body proportions, rigid surface appearance, joint morphology, and local texture, while robot-specific annotated data are often limited. This study presents a detection-guided framework for full-body 2D keypoint estimation from monocular video frames. A robot-specific detector first localises the target, and the detected box is expanded and normalised into a local region of interest (ROI) for keypoint recovery. Rather than treating the task as direct coordinate regression, the proposed integration uses ResNet18 feature extraction, convolutional block attention module (CBAM) refinement, heatmap-based keypoint representation, differentiable spatial to numerical transform (DSNT)-based continuous coordinate decoding, and skeleton-aware regularisation. A compact 13-keypoint annotation protocol is defined to describe the head, shoulders, elbows, hands, hips, knees, and feet. On the held-out test set, the proposed model reduces mean per-joint position error (MPJPE) from 27.98 px to 15.49 px relative to the ResNet18 direct-regression baseline and improves object keypoint similarity (OKS) from 0.80 to 0.96. Under the same 13-keypoint protocol, the proposed model also achieves a lower MPJPE than fine-tuned YOLOv8-Pose with detector-ROI input (18.85 px) and HRNet-W18 (30.82 px). Direct transfer of COCO-pretrained YOLOv8-Pose performs substantially worse, with an MPJPE of 201.50 px and an OKS of 0.28. These results support the effectiveness of robot-specific local spatial modelling and continuous coordinate recovery for monocular humanoid robot keypoint estimation under the evaluated limited-data setting. Full article

Camera-only BEV 3D object detectors are trained under highly imbalanced category distributions, and their matched object queries can exhibit directional semantic errors toward frequent classes. We investigate this behavior as a diagnostic problem: given fixed geometric predictions and fixed query–ground-truth assignments, how much class information remains accessible in frozen decoder features, which errors can be recovered, and where does recovery fail? We establish a scene-disjoint protocol in which recovery fitting and model selection use separate subsets of the official nuScenes training set, while all 150 validation scenes (6019 samples) remain final-only until all model and post-processing choices are fixed. Geometry-only Hungarian matching produces 158,253 fixed positive pairs on the full validation set. The frozen detector obtains a macro accuracy of 0.6136 on these pairs, while a lightweight factorized head trained on frozen features from decoder layer 4 reaches 0.6959 ± 0.0012 across three seeds. A linear probe achieves a macro accuracy of 0.8628 on the internal tuning split, whereas a shuffled-label control remains at chance (0.1000), indicating that substantial class information remains decodable from the frozen features. Tail-focused analysis further shows that recovered errors are more separable in frozen feature space than unrecovered errors across all 15 class-by-seed comparisons. However, recovery is not consistently observed across the controlled ResNet-18 and ResNet-50 configurations, and locked end-to-end evaluation decreases mAP from 0.2565 to 0.1620 and NDS from 0.3582 to 0.2796. These results support a geometry-controlled diagnosis of partial and class-dependent semantic recoverability, rather than improved localization, architecture-independent recovery, or deployable detection performance. Full article

Electromagnetic scattering from plunging breaking waves generates strong sea-surface radar returns that degrade radar-based maritime surveillance and target detection performance. This study develops a deep learning framework for automatic classification of simulated plunging-breaker scattering conditions using high-range-resolution profile (HRRP) data. The electromagnetic scattering data are generated using a physics-based Capillary Wave Modification Facet Scattering Model (CWMFSM) combined with ray-tracing techniques. Eight simulated plunging-breaker scattering conditions are constructed by combining two wind speeds, 7 m/s and 10 m/s, with four temporal conditions, Δt1, Δt10, Δt14, and Δt16. A total of 8000 HRRP samples are generated, with 100 normalized range-cell features extracted from each sample. Two deep learning classifiers, an artificial neural network (ANN) and a one-dimensional residual convolutional neural network (1D ResNet CNN), are comparatively evaluated. The ANN achieves an overall classification accuracy of 96%, compared with 91% for the 1D ResNet CNN under the simulated dataset and adopted model configurations. Robustness analysis under controlled additive white Gaussian noise (AWGN) conditions further shows that classification performance decreases as the signal-to-noise ratio is reduced, while noise-augmented training improves the robustness of both classifiers. Overall, the results demonstrate the feasibility of HRRP-based deep learning for distinguishing simulated plunging-breaker scattering conditions from sea-surface radar returns, providing a basis for further investigation of sea-clutter characterization and maritime radar applications. Full article

Geographical authentication of medicinal plant materials remains challenging because multispectral variables are often highly collinear and sample grouping can complicate reliable model validation. Existing correlation-based feature-selection strategies also require careful adaptation to multiclass problems to avoid artificial ordering of class labels and information leakage during model development. Therefore, this study aimed to develop a compact and leakage-controlled multispectral learning framework for geographical-origin discrimination. This study analyzed 800 physical Pinellia ternata samples from Gansu Xihe, Sichuan Neijiang, Sichuan Chengdu, and Chongqing Dianjiang (200 samples per origin). Each physical sample was represented by 31 mean grayscale intensities calculated from Otsu-segmented multispectral regions of interest. A Pearson-correlation-guided deep multilayer perceptron (PCG-DeepMLP) was constructed by estimating one-vs-rest band relevance and inter-band redundancy only within the training data. The key methodological innovation is a unified multiclass-aware, relevance–redundancy spectral-learning framework in which class-specific one-vs-rest Pearson relevance is coupled with inter-band redundancy control and embedded within leakage-controlled grouped model development. By learning the spectral subset exclusively from each training partition before nonlinear classification, the framework produces compact and complementary multispectral signatures while preserving multiclass structure and strict independence of held-out groups. Model and feature-selection settings were chosen by three-fold grouped cross-validation within each training partition. PCG-DeepMLP retained 9–21 bands and achieved the highest mean accuracy (0.9812 ± 0.0135), macro-F1 (0.9812 ± 0.0135), Matthews correlation coefficient (MCC; 0.9752 ± 0.0179), and macro-AUC (0.9994 ± 0.0006) among seven models. Its macro-F1 was higher than that of 1D-CNN, 1D-ResNet, full-band MLP, PLS-DA, and random forest after Holm correction. Performance was estimated through a strict nested group-wise internal validation scheme, with every outer test fold remaining isolated from feature selection, preprocessing, and model optimization. These findings demonstrate that multiclass-aware relevance–redundancy learning can retain complementary Pinellia ternata origin-discriminative information in a compact and stable spectral representation, enabling accurate geographical-origin authentication while providing a principled basis for reduced-channel acquisition and future independent multi-batch validation. Full article

Integrating blockchain technology into the marketing, communication, and logistics processes of sustainable supply chains faces multidimensional barriers. This study uses a mixed-methods approach to identify, prioritize, and analyze the expert-perceived influence relationships among these barriers. Semi-structured interviews with 14 managers and experts were thematically analyzed in MAXQDA to derive the barrier criteria, weighted and structurally analyzed using the Best-Worst Method (BWM) and Fuzzy DEMATEL with judgments from 11 domain experts, re-derived from the raw questionnaire data. BWM identifies Legal and Regulatory Barriers (C1, w = 0.2225) and Economic Barriers (C5, w = 0.2110) as most critical, jointly accounting for 43.4% of the total weight; nine of eleven experts’ solutions met the CR < 0.10 threshold; the remaining two (E4 and E9) were retained rather than adjusted. Fuzzy DEMATEL classifies Economic, Legal, and Trust barriers as cause-group factors and Technological, Institutional, and Knowledge barriers as effect-group factors, with Economic Barriers showing the strongest net driving influence (D − R = +0.705). The two highest-priority barriers are also the two most upstream, and sensitivity analysis confirms the top and bottom ranks are stable. Interventions targeting cause-group barriers may generate broader system-level effects across effect-group barriers, an expectation grounded in expert-perceived influence rather than demonstrated causal evidence. Full article

Accurate and subject-independent recognition of shoulder physiotherapy exercises from wrist-worn inertial measurement unit (IMU) signals is essential for automated home-based rehabilitation monitoring, yet existing deep learning models are too parameter-intensive to deploy on resource-constrained smartwatch hardware. This paper presents ECA-ResNet1D-Lite, a lightweight one-dimensional depthwise separable residual network augmented with efficient channel attention (ECA), trained and evaluated on the SPARS9x dataset comprising six shoulder exercises recorded from 20 subjects using a commercial wrist-worn smartwatch at 50 Hz. Because 50% window overlap allows adjacent windows to share samples, we report three protocols—window-level five-fold, recording-level grouped five-fold, and leave-one-subject-out (LOSO) cross-validation—with model selection performed throughout on validation data disjoint from the test partition. Under LOSO, the primary protocol, the model attains 99.1 ± 1.0% accuracy while requiring only 13,612 parameters and 0.46 M multiply–accumulate operations per window—the highest accuracy and the lowest between-subject dispersion of the seven architectures trained under an identical protocol, ahead of the strongest unconstrained baseline (InceptionTime, 98.8 ± 1.4%) at 36.3× fewer parameters and 213× fewer operations, and ahead of the parameter-efficient designs TinyHAR (97.4 ± 2.5%) and TinierHAR (97.4 ± 2.1%). An ablation over four attention variants (SE, ECA, CBAM, and multi-head self-attention) shows ECA to be the cheapest, adding six parameters (0.04% overhead), while delivering the largest LOSO gain over the same backbone without attention (+0.2 percentage points) and reducing the cross-subject standard deviation from 1.4% to 1.0%. Per-class analysis further reveals that shoulder girdle stabilization is the most challenging exercise under LOSO (F1-score: 96.4%), despite being the most represented class, attributable to its quasi-static, low-amplitude IMU signature. Deployed to an Apple Watch Ultra 2, the model classifies a 4-s window in 0.24 ms (duty cycle 0.012%), establishing that subject-independent shoulder physiotherapy monitoring is computationally feasible on current smartwatch hardware. Full article

This paper presents a computational pipeline for window-level semantic enrichment of texture-mapped building models used in urban digital twins (UDTs). The pipeline combines SAM-based candidate generation, super-resolution-based input matching (SRIM), a fine-tuned ResNet-50 window/non-window classifier, and texture-to-mesh mapping to instantiate verified regions as independent 3D window objects. Rather than proposing a new segmentation or super-resolution model, the study integrates existing components for low-resolution facade textures attached to 3D models. In experiments on 130 real building models, the EDSR-based SRIM configuration achieved the best mean accuracy of 95.0% and F1 score of 0.951 over 10 runs. An auxiliary experiment on cropped Open Images samples showed a consistent advantage of SRIM-based conditioning. A small-scale aspect-ratio-based evaluation of the generated 3D windows yielded an overall mean relative error of 14.93%, indicating suitability for semantic enrichment rather than precision-grade reconstruction. The method is relevant to downstream UDT applications such as facade editing, maintenance planning, and simulation-oriented model refinement. Full article

Accurate classification of lymphoma cell subtypes is essential for disease diagnosis and therapeutic decision-making, yet conventional approaches often rely on fluorescence labeling, labor-intensive sample preparation, and specialized instrumentation, limiting their applicability for rapid, label-free single-cell analysis. Here, we present an AI-assisted microfluidic light-scattering imaging platform for label-free classification of lymphoma cells. The platform integrates hydrodynamic focusing within a microfluidic chip, continuous acquisition of two-dimensional (2D) light-scattering patterns, automated image preprocessing, and transfer learning based on a pretrained ResNet50 network for intelligent optical feature extraction and classification. Human B lymphoma (Daudi) and T lymphoblastic lymphoma (SUP-T1) cells were used to evaluate the proposed framework. The optical imaging system was first validated using standard microspheres, demonstrating reliable acquisition of light-scattering patterns under continuous-flow conditions. A dataset comprising 800 single-cell scattering patterns was subsequently established and evaluated using stratified five-fold cross-validation. The proposed framework achieved an average classification accuracy of 94.75% with an average area under the receiver operating characteristic (ROC) curve of 0.986. By integrating microfluidic optical biosensing with deep learning, this work enables automated interpretation of intrinsic optical scattering signatures and provides a promising AI-enabled strategy for rapid, label-free lymphoma screening and intelligent healthcare applications. Full article

The shoe last is central to footwear design. Modern footwear CAD operates on parametric digital lasts, yet much last geometry—legacy collections and lasts that skilled last makers still sculpt by hand and copy by pantograph turning—exists only as physical models or as point-cloud scans lacking the structured parametric form that footwear CAD requires. This paper presents a complete template-based method for reconstructing a watertight parametric last from a segmented point cloud without intermediate triangulation. The only manual input is three landmark points—for which the system proposes standard positions—and the interactive confirmation of two boundary lines on the digitized last. From these, the method defines four feature points, a median plane, and a four-curve boundary network; all subsequent stages run without user interaction. A curvature-adaptive quadrilateral grid is constructed on the cloud by geodesic tracing and monitor-weighted area-orthogonality relaxation. A periodic Coons tube interpolates the grid and initializes the parameterization for a periodic tensor-product cubic B-spline surface fitted by penalized least squares with cyclic/open difference penalties, exact boundary interpolation, and toe-aware weighting. Cap surfaces close both collar and sole openings, and the model is exported as a watertight B-rep solid. Tests on sixteen industrial lasts using one fixed parameter set produced a mean one-sided deviation of 0.034 mm (RMS 0.058 mm) from the withheld industrial reference meshes in approximately 12 s per last. With synthetic noise at 50 dB SNR, the mean deviation increased by only 0.011 mm. A sampling-density study indicated near-second-order convergence before the control-net reaches an upper limit. The resulting solids import directly into CAD systems and support re-lasting, footwear design, and customization. Full article

Intelligent sheep behavior monitoring is vital for modern husbandry, but faces severe challenges in natural pastures due to high-density flock occlusion. Traditional 2D skeleton-based networks often suffer from depth ambiguity and feature collapse, misclassifying static tremors as dynamic displacement. To overcome this, we propose a robust multimodal dual-stream framework using skeletal and local visual fusion. The architecture features an upstream spatial perception stage utilizing YOLOv11m-Pose. To reduce annotation costs and improve robustness, we introduce an Active Hard-Example Mining mechanism, explicitly retaining difficult samples with severe overlapping or edge truncation. For downstream behavioral decisions, a multimodal dual-stream architecture processes the targets. The Spatio–Temporal Kinematic Stream employs a Kinematic Denoising Engine, incorporating a 1D Gaussian filter and displacement dead-zone gate to purify 2D coordinates before feeding them into a BiLSTM network. Concurrently, the Spatial Visual Stream uses a ResNet-50 backbone on cropped RGB patches to capture essential spatial context, addressing the limitations of pure coordinates. Finally, a weighted Softmax layer integrates both streams. Experiments on a complex real-world dataset validate this approach. A baseline kinematic-only model achieved just 69.05% overall accuracy and 68.18% walking precision. In contrast, our dual-stream fusion network achieved 93.26% overall accuracy, elevating walking precision to 97.14% and the eating F1-score to 94.29%. By effectively decoupling similar static and dynamic behaviors, this study demonstrates the indispensability of local visual features, establishing a high-precision baseline for smart livestock monitoring. Full article

Background/Objectives: Tetralogy of Fallot (TOF) is the most common cyanotic congenital heart defect, and cardiac computed tomography (CT) is increasingly central to its anatomical and pre-procedural assessment. Artificial-intelligence research in TOF is dominated by MRI; deep learning on cardiac CT in congenital heart disease exists but addresses multi-class diagnosis and segmentation, and the one binary TOF-versus-control CT study used slice-level validation without confounder control, and, to our knowledge, no CT study reports controlling the confounding intrinsic to a TOF-versus-control comparison. This confounding is structural: TOF is imaged predominantly in infancy, so a naive classifier can learn age, body size, and acquisition protocol rather than pathology. We develop and internally evaluate a confounder-matched, anatomy-guided deep-learning pipeline for TOF on cardiac CT. Methods: Contrast-enhanced cardiac CT from a single scanner was de-identified and restricted to one reconstruction (FC15 kernel, 0.5 mm), then matched 1:1 on age and sex, yielding 42 TOF and 42 controls (n = 84); controls were children imaged for suspected but excluded cardiovascular disease, so scan indication, unlike age and sex, was not matched. Standardized volumes were decomposed into four fixed sub-volumes positioned to approximate the components of the diagnostic tetrad: malalignment ventricular septal defect (VSD), overriding aorta, right-ventricular outflow tract (RVOT), and right-ventricular hypertrophy (RVH). Whether each sub-volume contains its named target was audited against independent physician region-of-interest annotations. Per region, a 2.5D transfer-learning classifier (ImageNet ResNet18) and a 3D CNN (DenseNet121) were trained with leak-free patient-level five-fold cross-validation and the branches fused. Optimism was assessed by repeated cross-validation and, for model selection, by nested cross-validation with the component subset and operating point chosen inside an inner loop. Discrimination was reported with bootstrap 95% confidence intervals (CIs); AUROCs were compared by DeLong test, with Benjamini–Hochberg correction applied to a seven-member family (the four within-component comparisons, two hybrid-versus-VSD contrasts, and hybrid versus whole-heart) and other comparisons reported uncorrected. Results: Matching removed the age difference (median 0.33 years, IQR 0.17–0.92 vs. 0.33, IQR 0.27–0.73; p = 0.86) with balanced sex (p = 1.00). The pre-specified four-component hybrid reached AUROC 0.829 (95% CI 0.74–0.91); the VSD region alone reached 0.828 (0.74–0.91), so the tetrad decomposition did not improve accuracy, and the containment audit shows it does not deliver the intended anatomical interpretability either. The 2.5D model exceeded the 3D CNN for every component (0.769–0.828 vs. 0.573–0.656; raw DeLong p = 0.007–0.037, Benjamini–Hochberg q up to 0.065 under a seven-member family, the weakest comparison (RVH) not surviving correction). Repeated cross-validation gave 0.811 ± 0.026 and nested cross-validation 0.787 ± 0.029; a stronger backbone with multi-phase data, handcrafted radiomics, and a large CT foundation model did not significantly improve on the matched pipeline. Grad-CAM maps were sensitive to both model weights and labels and superior to a centred-blob null in all eight comparisons and significantly so in seven, but not consistently superior to a resolution-matched random attribution, so no localization claim is made. Calibration was imperfect (slope 0.67) and recalibration gave no net gain; at an in-sample Youden threshold sensitivity was 0.93 and specificity 0.64. Occlusion sensitivity on the whole-heart baseline model showed it relies on the physician-marked septal, aortic and right-ventricular sites 1.8–4.1 times more than distance-matched surrounding tissue, while gross morphometry alone reached 0.651–0.663. Two of the four sub-volumes did not contain their target: the RVOT prior contained the physician annotation in 48.1% of cases and, because the model samples only the central band, excluded it in 99.4%; the RVH box was offset toward the midline, containing the marked target in 43.4% of annotations. Repositioning the priors, leak-free and derived from controls only, did not change discrimination (all p ≥ 0.10), and boxes placed at random positions inside the standardized heart reached 0.765 on average against 0.796 for the published priors, a difference this cohort cannot resolve. Conclusions: As a proof of concept, confounder-matched deep learning can recognize TOF on cardiac CT. Increasing model capacity did not significantly improve on the matched pipeline; the separate contribution of matching itself was not isolated against an unmatched comparator. Given the small, single-centre sample and the absence of external validation, these findings are hypothesis-generating and require external, multi-centre confirmation before any clinical use. Full article

Granulation is a major physiological disorder that compromises sweet orange quality. Conventional destructive detection methods lead to food waste and cannot be used for batch inspection. This study proposes a nondestructive approach for detecting granulation in sweet oranges using air-jet transient excitation and a laser Doppler vibrometer (LDV). Acoustic vibration spectra were acquired from 640 sweet orange samples. Using both competitive adaptive reweighted sampling (CARS)-extracted feature parameters and raw acoustic vibration spectra as inputs, an ISNet-1D model integrating a multi-scale Inception module and a squeeze-and-excitation (SE) attention mechanism was developed, and its performance was compared against those of partial least squares discriminant analysis (PLS-DA), support vector machine (SVM), K-nearest neighbor (KNN), random forest (RF), and baseline deep learning models including one-dimensional convolutional neural network (1D-CNN), Visual Geometry Group network 16 (VGG16), and residual network v1 (ResNet-v1). The results demonstrated that the ISNet-1D model trained on the full raw vibration spectrum achieved the best performance, with an overall test set accuracy, recall, and specificity of 92.97%, 95.00%, and 91.18%, respectively. Ablation experiments revealed that removal of the Inception branches and the SE module reduced the overall test accuracy by 5.47% and 4.69%, respectively, indicating that their combination effectively extracts multi-scale acoustic vibration features and enhances model precision. Gradient-weighted class activation mapping further identified the critical frequency bands primarily relied upon by the model for prediction. Collectively, noncontact acoustic vibration detection combined with ISNet-1D provides a viable method for nondestructive granulation detection in sweet oranges. Full article

Classifying benign and malignant lung nodules from computed tomography (CT) images remains difficult because lung nodules can be hard to classify, and unimodal models cannot capture complementary diagnostic information. Despite the success of deep learning, existing methods rely only on image information and miss semantic information that can be obtained from an expert radiologist’s knowledge. Hence, the authors propose a new multimodal lung nodule classification model in this study, named the Graph-Guided Fusion Network (R-GGFN), that combines three-dimensional (3D) CT image features and structured radiologist annotations. The proposed architecture consists of three models. A 3D ResNet-18 network for image feature extraction, an MLP network for encoding semantic information, and a Graph Attention Network (GAT) for capturing inter-nodule relationships and fusing multimodal information with the graph. We add a tabular skip connection to preserve discriminative semantic features and use focal loss to address imbalance during training. To prevent data leakage, we partitioned the publicly available LIDC-IDRI dataset at the patient level. Experimental results on a held-out patient-level test set, accessed only once after model selection was finalized, show that the proposed R-GGFN achieves an accuracy of 85.21%, an AUROC of 0.9147, a PR-AUC of 0.9213, and an F1-score of 0.8609. Among all unimodal and multimodal baselines internally evaluated, R-GGFN achieved the best value on every reported metric, including accuracy, AUROC, PR-AUC, F1-score, Precision, sensitivity, and specificity. Furthermore, the proposed approach enhances model transparency by combining explainable AI techniques (e.g., 3D Grad-CAM, SHAP, and graph visualization) to explain the model at the image, feature, and graph levels. The results show that the graph-guided multimodal fusion method can fully leverage complementary image and semantic information, improving diagnostic accuracy and interpretability. The framework proposed here is a good and understandable computer-aided diagnosis decision-support system for lung cancer and a step towards future external dataset validation. Full article

Multi-camera vehicle sensing provides a cost-effective solution for 3D environmental perception in intelligent vehicles. A practical challenge is that temporal fusion becomes unreliable when the ego-vehicle undergoes aggressive motion. Historical camera features are commonly aligned by rigid ego-pose warping, but residual motion-induced displacement can still degrade object localization, heading estimation, and velocity sensing. This paper presents GCL-BEV, a motion-aware temporal compensation framework for multi-camera vehicle sensor systems. The proposed framework uses synchronized surround-view cameras and ego-motion measurements as coupled sensing inputs. First, a Geometric-Aware Feature Enhancement (GAFE) module converts ego-motion priors into motion-conditioned BEV sampling offsets, allowing the visual sensing representation to compensate for local temporal misalignment before fusion. Second, a View-Consistency Learning (VCL) objective imposes a training-time equivariance constraint so that the sensor representation remains consistent under planar viewpoint perturbations. Across 10 random seeds on nuScenes, GCL-BEV achieves 57.80% ± 0.15 NDS and 46.22% ± 0.16 mAP with a ResNet-101 backbone. Compared with BEVDet4D, it reduces the mean Average Orientation Error by 5.4% and shows smaller degradation from steady driving to high-turn scenarios, indicating improved robustness for dynamic vehicle sensing. Full article

Cross-condition fault identification for crane slewing bearings is difficult because low rotational speed, heavy loading, and operating-condition variation jointly weaken fault-induced impulses and alter the distribution of monitoring data. In addition, vibration and acoustic emission signals describe different aspects of bearing degradation, making fixed multi-sensor fusion insufficient when sensor sensitivity changes across fault states. This study proposes a Lightweight Domain-Adaptive Graph Convolutional Network (LDAGCN) for cross-condition diagnosis. The model uses six vibration channels and one acoustic emission channel as synchronized heterogeneous inputs. Two compact one-dimensional encoders first learn modality-specific temporal representations, after which a feature-wise gate determines the relative contribution of each sensing modality. The fused embeddings in every mini-batch are regarded as graph nodes, and a sparse sample graph is reconstructed from Top-k cosine similarities. Multi-receptive-field graph convolution then aggregates one-hop and higher-order neighborhood information, while a residual connection limits excessive modification of the original fused features. To reduce the discrepancy between operating conditions, the training objective combines source-domain classification, adversarial domain discrimination, maximum mean discrepancy, and supervision from a small labeled target-domain adaptation subset. Experiments were carried out on a dedicated crane slewing-bearing test rig containing normal, inner-race fault, outer-race fault, and B1 localized-fault states. On the target-condition test set, LDAGCN achieved an accuracy of 97.22%, a Macro-F1 score of 97.21%, a Macro-Precision of 97.37%, and a Macro-Recall of 97.22%. The proposed model also outperformed 1D-CNN, ResNet1D, CNN-LSTM, DANN, MMD-DAN, and DeepCORAL under the same data partition. The confusion matrix and t-SNE projection indicate that the learned representation reduces the source–target distribution gap while retaining fault-class separation. The ablation results further indicate that the acoustic emission information, gated fusion, graph-based association learning, and domain adaptation complement each other in terms of the final diagnostic performance, while maintaining a lightweight architecture that is suitable for practical monitoring. Full article