Search Results (614)

Under the dwarf-rootstock dense planting method, existing apple detection models for intelligent harvesting suffer from excessive parameter counts that hinder deployment on resource-constrained devices, while lightweight alternatives often sacrifice detection accuracy. To address this dilemma, this paper proposes YOLO-SPM, a lightweight apple detection model based on the YOLOv12n architecture, specifically designed for complex orchard environments. The core innovation lies in a problem-driven, three-stage collaborative optimization strategy: first, PConv is introduced to replace standard convolutions in the A2C2f module, reducing computational redundancy by exploiting channel-wise feature similarity of apple targets; second, the parameter-free SimAM attention mechanism is embedded in the neck network to enhance the model’s focus on occluded fruit features without increasing model size, while MBConv is integrated into the detection head to further reduce computational cost; third, WIoU v3 is adopted as the loss function to compensate for the accuracy loss incurred by lightweight design through its dynamic focusing mechanism on difficult samples. This complementary design ensures that each module addresses a distinct bottleneck of the native YOLOv12n in orchard scenarios, achieving a balance between efficiency and accuracy rather than simple module stacking. Experimental results demonstrate that YOLO-SPM achieves a precision of 92.8% and mAP@0.5 of 93.1%, outperforming the baseline by 4.8 and 5.3 percentage points, respectively, while reducing parameter count, FLOPs, and memory footprint by 40.2%, 35.4%, and 41.8%. This study provides a feasible solution for high-precision apple identification in dwarf-rootstock dense planting orchard environments, with the potential for integration into automated harvesting systems upon future on-device validation. Full article

Bolting time is a pivotal agronomic trait that determines the yield and commercial quality of Brassica rapa ssp. chinensis var. utilis. To investigate the molecular basis of early bolting, an early-bolting line ‘m662’ and a late-bolting line ‘t151’ were used in this study. Phenotypic evaluation combined with shoot apical meristem (SAM) observation showed that 10 days of low-temperature vernalization markedly accelerated bolting in ‘t151’. Subsequently, SAM samples from ‘m662’, non-vernalized ‘t151’, and 10-day vernalized ‘V10-t151’ were collected at five developmental stages (7, 10, 13, 16, and 19 d after transplanting) for transcriptome sequencing. Weighted gene co-expression network analysis revealed that key module genes related to gibberellin signaling were specifically enriched in ‘m662’ before bolting, whereas those in the middle and late bolting stages were enriched in hormone response, cell cycle regulation, and floral organ development. In ‘t151’, hub genes detected at 7–13 d included three paralogs of the floral integrator gene SOC1 and BraA06.FPF1. BrSOC1 (BraA03g024230.4C) was significantly upregulated in response to vernalization. DEGs identified during the late developmental stage (16–19 d) included genes involved in transmembrane transport processes, flower development, reproductive shoot system development. Expression analysis across the three materials showed that vernalization accelerated bolting in ‘t151’ by repressing BrFLC expression and promoting BrSOC1 expression. This study elucidates the dynamic transcriptomic network underlying early bolting in non-heading Chinese cabbage, providing key functional genes and mechanistic insights for bolting regulation and molecular breeding. Full article

Beat-level synchrony among cardiomyocyte clusters is a critical indicator of cardiac electromechanical function. Traditional invasive approaches have substantial limitations, and conventional computer vision methods are poorly suited for resolving densely packed, adherent clusters. To address these challenges, we developed an analysis framework to characterize the beating characteristics of cardiomyocyte clusters from microscopic imaging data. Specifically, we propose CardioSegNet, a multi-task deep learning model that combines attention mechanisms with three prediction heads (semantic segmentation, contour detection, and distance transform), followed by a watershed algorithm to achieve high-accuracy cluster-level segmentation of cardiomyocyte clusters. The Pixel-Difference method is applied to extract time-series beating signals from each segmented cluster and compute several dynamic parameters, including beating amplitude, period, frequency, and the Beat Rate Irregularity (BRI). We further introduce PeriodAwareNAPTD_ij to quantify the beating synchrony among different clusters. Our experimental results show that CardioSegNet achieves a Dice coefficient of 0.8868 and an HD95 of 93.02 µm on an independent test set, demonstrating strong segmentation performance. The cardiomyocyte populations are not uniformly globally synchronized; rather, they consist of multiple local subgroups with high internal synchrony, and the degree of synchronization between clusters is positively correlated with their physical distance. This label-free analytical pipeline provides an efficient tool for myocardial function evaluation and cardiotoxicity screening in vitro. Full article

Rice leaf diseases severely threaten global food security, and efficient on-site detection remains challenging for resource-constrained field inspection robots. This work introduces a lightweight visual detection framework designed for the real-time and accurate identification of rice leaf diseases on agricultural edge mobile platforms. A dataset of 4622 annotated images compiled from mobile-device acquisition and publicly available online sources, covering three representative disease categories, together with an independent public benchmark, was used for evaluation. The framework integrates three complementary modules: adaptive multi-scale feature extraction via a dynamic hybrid convolution backbone (C3k2-DICN), cross-scale parameter sharing in the detection head (CSDH) to reduce redundancy, and dual-path downsampling (ADown) to preserve disease-discriminative information during resolution compression. Compared to the YOLO11n baseline, the proposed approach reduced GFLOPs by 36.5% and parameter count by 34.6%, while achieving 88.42% mAP@0.5 and 45.82% mAP@0.5:0.95 on the compiled dataset and 91.71% mAP@0.5 on the public benchmark, indicating accuracy competitive with or superior to all evaluated comparison models. Deployed on an NVIDIA Jetson TX2 with TensorRT FP16 acceleration, the model ran in real time on-device, reaching 32.2 FPS for the TensorRT inference stage and 19.8 FPS for the full end-to-end pipeline including image pre- and post-processing. The framework offers a practical basis for lightweight on-device rice disease detection; closed-loop validation on a moving field robot is left to future work. Full article

Accurate and rapid detection of Volvariella volvacea (straw mushroom) fruiting bodies at harvestable maturity is a critical prerequisite for automated industrial cultivation. However, existing detection methods often yield high false-negative and false-positive rates when processing a small-scale, densely distributed, and heavily occluded targets against complex straw substrate backgrounds. Furthermore, these methods frequently struggle to balance the competing requirements of architectural efficiency (such as parameter volume and computational complexity) and real-time performance for edge computing. To address these challenges, this study proposes a YOLO11n-CPDM, a lightweight detection model based on an improved YOLO11n architecture. The model incorporates synergistic optimizations across feature extraction, fusion, and reconstruction. First, a Dual Coordinate Attention Feature Extraction mechanism is integrated into the C3k2 bottleneck blocks of the backbone network. This enhances target perception in complex, occluded environments by concurrently modeling global context and local salient features. Second, within the neck network, the standard attention module is replaced with the PnPNystraAttention module, coupled with the DySample dynamic upsampling operator. This modification strengthens contextual relationships among multi-scale features and improves spatial consistency during reconstruction while preserving linear computational complexity. Finally, the detection head is optimized using MBConv blocks based on an inverted residual structure to minimize parameter volume. Experimental results on a custom V. volvacea dataset demonstrate that the proposed YOLO11n-CPDM model achieves significant performance gains, with Precision (P), Recall (R), and Mean Average Precision (mAP50) reaching 86.8%, 87.5%, and 88.4%, respectively. These figures represent improvements of 2.7, 3.0, and 3.2 percentage points over the baseline YOLO11n model. Additionally, the model size is reduced to 4.8 MB (a 12.7% decrease), while achieving inference speeds of 42.7 FPS on Jetson AGX Orin and 21.2 FPS on Jetson Nano, outperforming the baseline model on both embedded platforms. Consequently, the proposed model effectively enhances detection performance in complex environments while maintaining excellent lightweight characteristics and deployment flexibility, providing a solid technical foundation for intelligent perception and automated harvesting of V. volvacea. Full article

Detecting objects in drone-captured aerial imagery is particularly formidable due to challenges such as the prevalence of numerous small targets and their dense spatial distribution. To bridge this gap, this paper introduces YOLO-UTD (YOLO-UAV Traffic Detection), a dedicated small object detector tailored for drone traffic surveillance. Built upon the YOLOv8 framework, the proposed model incorporates three principal enhancements. First, a specialized small-object detection head replaces the original large-object head to increase the sensitivity to fine-grained features. Second, we introduce a shallow-augmented feature pyramid network (SFPN) into the neck module. The SFPN enriches the semantic content of high-resolution shallow features via dense multiscale interactions and CARAFE upsampling, boosting performance on small targets. Finally, a C2fA layer is integrated into the deep backbone stages to adaptively fuse spatial details and semantic context through a dual-path architecture and a cross-attention mechanism, thereby dynamically refining features critical for small objects. Extensive experiments on the VisDrone2019 dataset validate that YOLO-UTD achieves a 3.6% higher mean average precision (mAP) than YOLOv8 while preserving a low parameter footprint, with a particularly significant gain of 5.3% in vehicle detection accuracy. These findings confirm the model’s efficacy and strong potential for application in smart city drone surveillance. Full article

Unmanned Aerial Vehicle (UAV) image object detection holds significant application value in the low-altitude economy, traffic monitoring, intelligent agriculture, and disaster rescue. However, due to the top-down perspective, UAV images typically suffer from challenges such as small target scales, dense object distribution, severe occlusions, and complex backgrounds. These issues often limit the recall and localization accuracy of general-purpose detectors when they are directly applied to UAV small-object detection scenarios. To address these aforementioned challenges, this paper proposes an Adaptive Dynamic Aggregation YOLO network, termed ADA-YOLO. The novelty of ADA-YOLO lies in its highly efficient combinatorial design specifically tailored for UAV small object detection, while retaining the efficient backbone of YOLOv8, we systematically reconstruct the neck and detection head to improve accuracy. Specifically, a high-resolution P2 detection branch is incorporated to construct a P2–P5 multi-scale prediction structure. Furthermore, the lightweight DySample dynamic upsampling module is adopted to replace traditional upsampling methods, and an Adaptive Spatial Feature Fusion (ASFF) mechanism is introduced to alleviate semantic conflicts and noise interference during multi-scale feature fusion. This synergistic combination explicitly addresses multi-scale representation challenges and enhances small-object detection performance in complex scenes. Comparative experiments with the baseline YOLOv8n on the VisDrone2019 dataset demonstrate that ADA-YOLO achieves an improvement of 11.3% in mAP@0.5 and 8.2% in mAP@0.5:0.95. The improved model achieves these performance gains with a modest parameter increase and acceptable computational complexity. Finally, ablation experiments further validate the effectiveness of each individual module and their synergistic gains. Full article

With the widespread adoption of intelligent terminals, edge devices, and distributed information systems in the financial domain, financial security sensing data exhibit multisource heterogeneity, dynamic temporal patterns, and high privacy sensitivity. Traditional centralized anomaly detection methods are no longer able to simultaneously satisfy the requirements of cross-institutional or cross-node collaborative modeling, client data privacy protection, and robust monitoring of transaction and system anomalies. To address this challenge, a data-local federated deep anomaly detection framework has been proposed for distributed financial security sensing systems. Initially, a local deep financial security sensing representation module is constructed to perform temporal encoding and attention-based modeling on multisource financial signals, including terminal operation status, network transaction communication, backend server operation, identity authentication, and anomaly alerts, thereby extracting representations relevant to anomalous behaviors. Subsequently, a data-local federated optimization and personalized aggregation mechanism is developed to enable cross-node knowledge sharing without transmitting raw transaction or client data, while local personalized detection heads are employed to adapt to non-independent and identically distributed (non-IID) financial institution data. Furthermore, an adversarially robust security detection and trust-aware aggregation strategy is introduced to enhance model stability under input noise, feature masking, anomaly camouflage, and potential malicious client updates. Experimental results demonstrate that the proposed method achieves an Accuracy of $92.37\%$, a Precision of $89.41\%$, a Recall of $88.26\%$, an F1-score of $88.83\%$, an AUC of $93.06\%$, and a PR-AUC of $89.15\%$ in the primary financial anomaly detection task, significantly outperforming baseline methods such as Isolation Forest, Autoencoder, LSTM, Transformer, FedAvg, FedProx, SCAFFOLD, and MOON. In robustness experiments, the method attains F1-scores of $87.95\%$, $86.42\%$, $86.88\%$, $84.57\%$, $86.73\%$, and $83.91\%$ under Gaussian noise, feature masking, temporal shift, adversarial perturbation, and $20\%$ and $30\%$ malicious client scenarios, respectively. Ablation studies further confirm the effectiveness of local representation learning, personalized federated optimization, adversarial training, and trust-aware aggregation mechanisms. Overall, the proposed approach provides an efficient intelligent anomaly detection solution for financial AI security monitoring scenarios characterized by data localization requirements, node heterogeneity, and attack perturbations. Full article

To address the challenges of detecting sugar orange fruits in complex natural orchard environments—where fruits are frequently occluded by leaves and branches and may be mutually occluded due to dense growth, leading to missed detections, false positives, and low detection confidence—we propose an improved algorithm based on YOLOv8n, named YOLO-FSEP. A Spatial-Channel Synergistic Attention (SCSA) module is introduced into the main network to enhance feature extraction capabilities; the IoU loss function is replaced with Focal_SIOU to improve the detection accuracy for difficult samples; and an SE attention mechanism is embedded in the detection head, with the addition of a P6 high-resolution detection layer to optimize multi-scale object performance. Experimental results on a self-built sugar orange dataset show that, compared to the baseline YOLOv8n, the improved model achieves a 0.9% increase in accuracy, a 1.3% increase in recall, and a 3.2% increase in mAP50-95, while maintaining an inference speed of 62.6 FPS. To evaluate the model under dynamic conditions, we performed a 200-frame continuous test of the 3D localization pipeline on a laptop with a RealSense D435i camera. The average YOLO inference time was 49.90 ms, post-processing (depth extraction and 3D coordinate conversion) took 0.24 ms, and the total processing time was 50.15 ms. Given that the typical response time for a robotic arm’s single positioning operation is 100–200 ms, this real-time performance meets the dynamic localization requirements of sugar orange harvesting. Full article

Accurate traffic sign detection is important for the safety of autonomous driving systems. However, fully supervised methods require a large amount of manual annotation, which is cost-prohibitive and time-consuming. Semi-supervised methods employ a small amount of labeled data and a large amount of unlabeled data to train the models, hence largely reducing the annotation costs. However, these methods have the following challenges: (1) with an imbalanced long-tail class distribution of traffic signs, they tend to achieve poor performance on tail classes; (2) they often fail to detect small traffic signs. To solve these issues, we propose a Semi-Supervised Traffic Sign Detection method with Dynamic Pseudo-Label Selection and Gated Feature Fusion-based Proposal Refinement. Firstly, we design a Class Distribution-based Dynamic Pseudo-Label Selection module (CD-DPLS) to select pseudo-labels for different classes based on the class distribution information, which reduces the tendency to select more pseudo-labels from head classes instead of tail classes, thereby improving the tail class detection performance. Secondly, we employ a Gated Feature Fusion-based Proposal Refinement strategy (GFF-PR) to refine detection proposals by fusing different-scale features with a gating mechanism, which facilitates the detection of small traffic signs. In addition, we use an Adaptive-Weight Focal Loss (AWFL), with which the weight of each pseudo-label is determined by the ratio between its classification confidence and the corresponding class-specific classification-confidence threshold. Experiments on traffic sign datasets demonstrate that the proposed method outperforms state-of-the-art semi-supervised approaches, with mAP₅₀ scores of 10.8% and 34.9% using only 1% and 10% labeled data, respectively. Full article

Unmanned aerial vehicle (UAV) object detection plays an essential role in modern visual perception, but it remains challenging because UAV imagery typically contains extremely small, densely distributed objects embedded in complex backgrounds. Conventional detectors, including the recent YOLOv12, are prone to losing critical spatial details during downsampling and often exhibit task misalignment between classification and localization, particularly under severe scale variations. To address these problems, this study proposes SPTD-YOLO, a small-object-aware pyramidal and task-aligned dynamic detector. Specifically, a Small Object Enhanced Pyramid (SOEP) is developed by incorporating SPDConv and CSPOmniKernel to preserve and refine shallow, fine-grained features. In addition, a high-resolution P₂ detection layer is introduced to increase spatial grid density and strengthen the structural representation of tiny objects. Furthermore, a Task-Aligned Dynamic Detection Head (TADDH) is designed to decouple and coordinate classification and regression through dynamic convolution and a synergistic dual-gating mechanism. Experiments on VisDrone2019 show that SPTD-YOLO improves mAP@0.5 by 8.37% and mAP@0.5:0.95 by 5.11% over YOLOv12 while maintaining practical efficiency for UAV edge deployment. Full article

Background/Objectives: Early and accurate discrimination of neurological conditions, dementia, stroke and healthy aging, remains a critical clinical challenge. Electroencephalography (EEG) is a non-invasive measure of brain dynamics and entropy-based features obtained from multichannel EEG have shown strong discriminative ability. However, existing deep learning approaches do not sufficiently address the combined challenges of small clinical cohorts and high-dimensional entropy feature spaces. In this study, a novel architecture is proposed for multi-class neurological EEG classification under extreme small-sample conditions. Methods: A novel dual-branch Channel-wise Transformer and Attention-Branch Network (EEG-ChTABNet) are pr to classify 19-channel EEG entropy features into three classes (dementia, stroke, healthy control; N = 45; 15 per class). The architecture suggests four new designs. First, the Channel Importance Attention (CIA) block, which adaptively learns to re-weight the importance of electrodes via squeeze-excitation. Second, the dual-branch encoder, which combines the global multi-head self-attention with the local depthwise-separable convolution. Third, the gated sigmoid fusion mechanism. Fourth, the bottleneck residual classification head, to solve overfitting. Eight entropy feature sets: Amplitude-Aware Permutation Entropy (AAPE), Attention Entropy (AttEn), Dispersion Entropy (DisEn), Distribution Entropy (DistrEn), Fluctuation-based Dispersion Entropy (FDispEn), Fuzzy Entropy (FuzEn), Linear Gaussian Estimation of the Conditional Entropy (LinEn), and Symbolic Dynamics (SyDy) were evaluated individually with stratified 5-fold cross-validation on within-fold SMOTE augmentation. Results: EEG-ChTABNet consistently outperformed the baseline Transformer on all 8 feature sets. DisEn and SyDy features yielded peak classification accuracy of 73.3% (AUC: 0.823 and 0.857, respectively) compared to the corresponding baseline of 57.8% and 55.6%. SyDy achieved the best overall AUC of 0.857 and the dementia detection sensitivity was up to 86.7% over multiple feature sets. Conclusions: EEG-ChTABNet shows the effectiveness of channel-adaptive, dual-branch Transformer Designs for EEG-based neurological classification from Small-Sample Entropy Feature Data, and Identifying SyDy and DisEn as the Most Discriminative Feature Representations for Three-Class Neurological EEG Classification. Full article

High-frequency FY-4B aerosol optical depth (AOD) observations provide useful spatiotemporal constraints for dust nowcasting, but their application over bright deserts is limited by retrieval gaps and high-AOD uncertainty. This study develops a physics-informed spatiotemporal learning framework for 15–60 min FY-4B AOD nowcasting over the Taklimakan Desert. Historical FY-4B AOD, valid masks, ERA5 dynamic fields, model-level diagnostics, and surface constraints are organized on a unified 48 × 64 grid. An LSTM–TCN–Transformer temporal backbone is combined with spatial-context encoding, mask-aware observation encoding, and structured source–transport prediction heads to represent both temporal evolution and spatial plume structures. A physics encoder represents boundary-layer mixing, vertical wind shear, source-region emission, upwind transport, and deposition loss. Mask-aware encoding and structured prediction heads are used to handle missing retrievals, source and transport increments, high-AOD tails, and low-confidence regions. Results show that FY-4B AOD constrains the main dust-belt position and spatial extent within 1 h, with skill decreasing from 15 to 60 min. High-coverage samples show more stable spatial structures, whereas low-coverage and extreme high-AOD cases have larger peak underestimation and boundary errors. The proposed framework improves high-AOD event detection and spatial-structure preservation compared with persistence, advective persistence, ConvLSTM, and ST-UNet baselines. An additional case-based comparison with MODIS MAIAC AOD and MERRA-2 dust optical depth shows partial spatial colocation between predicted high-value footprints and independent aerosol-enhancement references; however, the reported skill scores should still be interpreted mainly as spatiotemporal consistency with the FY-4B AOD product field rather than direct validation of true atmospheric dust loading. Full article

Accurate wheat spike detection is essential for crop phenotyping and yield estimation, but real-world field conditions—such as dense spike overlap, environmental domain shifts, and degradation-induced failures like motion blur—pose significant challenges. Achieving robust perception under these circumstances while maintaining a strict accuracy-efficiency trade-off for edge devices remains a pressing research problem. To overcome these limitations, we propose GG-YOLO, a unified lightweight detection framework specifically tailored for complex agricultural environments. Rather than a simple recombination of existing lightweight modules, GG-YOLO integrates three original structural adaptations: First, a Dual-path Attentive Ghost Mechanism (DAGM) introduces gradient-guided attention modulation to enhance feature discrimination and explicitly resolve feature confusion in dense, overlapping regions. Second, a C3Ghost module combines multi-branch aggregation with linear feature generation, mitigating parameter redundancy in the prediction head by approximately 31% compared to the standard YOLOv8s without sacrificing semantic capacity. Third, DSample, a dynamic upsampling operator featuring an original dual-mode adaptive mechanism, robustly recovers fine-grained spatial details during multi-scale feature pyramid fusion. Extensive cross-dataset experiments on the GlobalWheat2020 and HNKJXYwheat datasets validate the model’s exceptional resilience to domain shifts and varying growth stages. GG-YOLO achieves a precision of 94.35%, a recall of 91.93%, and a state-of-the-art mAP@50 of 96.47%. Furthermore, the model contains only 7.89 M parameters and requires 20.4 GFLOPs, reaching an inference speed of 165 FPS on a desktop GPU and a validated real-time speed of 64 FPS on an NVIDIA Jetson edge computing platform. These results demonstrate that GG-YOLO establishes a superior accuracy-efficiency frontier, making it highly reliable for real-time field deployment in precision agriculture. Full article

Text detection models based on DBNet have demonstrated strong performance in natural scene text detection. However, these models still suffer from the following three issues. Firstly, the amplifying factor hyperparameter in the differentiable binarization (DB) makes it difficult for the text detection model to achieve optimal performance. Secondly, the integration of low-level and high-level features within the backbone’s feature pyramid lacks specific optimization strategies. Thirdly, the deconvolution operation in the prediction head may damage text contours. To tackle the aforementioned issues, this paper presents a text detection model termed ASM-DBNet, which mainly consists of three innovations. For the first issue, an adaptive differentiable binarization (ADB) scheme is proposed. It can independently predict amplifying factor for feature points at different spatial locations and replace the original amplifying factor hyperparameter, thereby improving the overall optimization performance of the model. For the second issue, a spatial-channel self-attention (SCA) module is proposed to optimize the fusion of high-level and low-level features. On the one hand, spatial self-attention is used to enhance the spatial localization ability of high-level features; on the other hand, channel self-attention based on a grouped transformer is used to optimize the fusion results of high-level and low-level features. For the third issue, a multi-scale context-enhanced dynamic upsampling (MC-DyUpS) module is proposed to replace the deconvolution operation in the prediction head. It enhances contextual perception in the region of interpolation points through multi-scale context feature extraction, and then accurately predicts coordinate offsets of interpolation points. The position correction based on these offsets effectively suppresses the spatial deviation caused by deconvolution. Ablation studies demonstrate the effectiveness of the SCA module, MC-DyUpS module, ADB scheme, and their arbitrary combinations. Comprehensive quantitative evaluations demonstrate that ASM-DBNet achieves competitive F1-scores of 84.1%, 84.2%, and 85.7% on the ICDAR 2015, Total-Text, and MSRA-TD500 datasets, respectively, with improvements of 1.8%, 1.4%, and 2.9% over the baseline model. Full article