It is challenging to accurately detect flexible objects with arbitrary orientation from monitoring images in power grid maintenance and inspection sites. This is because these images exhibit a significant imbalance between the foreground and background, which can lead to low detection accuracy when using a horizontal bounding box (HBB) as the detector in general object detection algorithms. Existing multi-oriented detection algorithms that use irregular polygons as the detector can improve accuracy to some extent, but their accuracy is limited due to boundary problems during the training process. This paper proposes a rotation-adaptive YOLOv5 (R_YOLOv5) with a rotated bounding box (RBB) to detect flexible objects with arbitrary orientation, effectively addressing the above issues and achieving high accuracy. Firstly, a long-side representation method is used to add the degree of freedom (DOF) for bounding boxes, enabling accurate detection of flexible objects with large spans, deformable shapes, and small foreground-to-background ratios. Furthermore, the further boundary problem induced by the proposed bounding box strategy is overcome by using classification discretization and symmetric function mapping methods. Finally, the loss function is optimized to ensure training convergence for the new bounding box. To meet various practical requirements, we propose four models with different scales based on YOLOv5, namely R_YOLOv5s, R_YOLOv5m, R_YOLOv5l, and R_YOLOv5x. Experimental results demonstrate that these four models achieve mean average precision (mAP) values of 0.712, 0.731, 0.736, and 0.745 on the DOTA-v1.5 dataset and 0.579, 0.629, 0.689, and 0.713 on our self-built FO dataset, exhibiting higher recognition accuracy and a stronger generalization ability. Among them, R_YOLOv5x achieves a mAP that is about 6.84% higher than ReDet on the DOTAv-1.5 dataset and at least 2% higher than the original YOLOv5 model on the FO dataset.

Grounded Visual Question Answering systems place heavy reliance on substantial computational power and data resources in pretraining. In response to this challenge, this paper introduces the LCV2 modular approach, which utilizes a frozen large language model (LLM) to bridge the off-the-shelf generic visual question answering (VQA) module with a generic visual grounding (VG) module. It leverages the generalizable knowledge of these expert models, avoiding the need for any large-scale pretraining. Innovatively, within the LCV2 framework, question and predicted answer pairs are transformed into descriptive and referring captions, enhancing the clarity of the visual cues directed by the question text for the VG module’s grounding. This compensates for the limitations of missing intrinsic text–visual coupling in non-end-to-end frameworks. Comprehensive experiments on benchmark datasets, such as GQA, CLEVR, and VizWiz-VQA-Grounding, were conducted to evaluate the method’s performance and compare it with several baseline methods. In particular, it achieved an IoU F1 score of 59.6% on the GQA dataset and an IoU F1 score of 37.4% on the CLEVR dataset, surpassing some baseline results and demonstrating the LCV2’s competitive performance.

Object detection methods are commonly employed in power safety monitoring systems to detect violations in surveillance scenes. However, traditional object detection methods are ineffective for small objects that are similar to the background information in the power monitoring scene, which consequently affects the performance of violation behavior detection. This paper proposed a small object detection algorithm named HS-YOLO, based on High-Resolution Network (HRNet) and sub-pixel convolution. First, to fully extract the microfeature information of the object, a small object feature extraction backbone network is proposed based on the HRNet structure. The feature maps of different scales are processed by multiple parallel branches and fused with each other in the network. Then, to fully retain the effective features of small objects, the sub-pixel convolution module is incorporated as the upsampling operator in the feature fusion network. The low-resolution feature map is upsampled to a higher resolution by reorganizing pixel values and performing padding operations in this module. On our self-constructed power operation dataset, the HS-YOLO algorithm achieved a mAP of 87.2%, which is a 3.5% improvement compared to YOLOv5. Particularly, the dataset’s AP for detecting small objects such as cuffs, necklines, and safety belts is improved by 10.7%, 5.8%, and 4.4%, respectively. These results demonstrate the effectiveness of our proposed method in detecting small objects in power operation scenarios.

Electric power operation is necessary for the development of power grid companies, where the safety monitoring of electric power operation is difficult. Irregular deformable objects commonly used in electrical construction, such as safety belts and seines, have a dynamic geometric appearance which leads to the poor performance of traditional detection methods. This paper proposes an end-to-end instance segmentation method using the multi-instance relation weighting module for irregular deformable objects. To solve the problem of introducing redundant background information when using the horizontal rectangular box detector, the Mask Scoring R-CNN is used to perform pixel-level instance segmentation so that the bounding box can accurately surround the irregular objects. Considering that deformable objects in power operation workplaces often appear with construction personnel and the objects have an apparent correlation, a multi-instance relation weighting module is proposed to fuse the appearance features and geometric features of objects so that the relation features between objects are learned end-to-end to improve the segmentation effect of irregular objects. The segmentation mAP on the self-built dataset of irregular deformable objects for electric power operation workplaces reached up to 44.8%. With the same 100,000 training rounds, the bounding box mAP and segmentation mAP improved by 1.2% and 0.2%, respectively, compared with the MS R-CNN. Finally, in order to further verify the generalization performance and practicability of the proposed method, an intelligent monitoring system for the power operation scenes is designed to realize the actual deployment and application of the proposed method. Various tests show that the proposed method can segment irregular deformable objects well.

Deploying accurate defect detectors on resource-constrained edge devices requires balancing computational cost and feature representation quality. To address this challenge, we propose HEC-YOLO, a lightweight detector for tiny and low-contrast industrial surface defects. The backbone is redesigned with a Multi-Scale Convolution Block (MSCB), which combines grouped multi-receptive-field processing and channel shuffle for efficient scale-adaptive feature extraction, and an Efficient Coordinate Attention Block (ECAB), which reconstructs C2f-style aggregation through lightweight feature transformation and structural re-parameterization. Within ECAB, the proposed Global-Coordinate Synergistic Attention (GCSA) jointly models global channel dependencies and orthogonal coordinate-aware information to enhance defect localization. In the neck, a Context-Guided Feature Pyramid Network (CGFPN) integrates an efficient fusion topology, an SPD-Conv-based shallow-detail pathway, and complementary inter-layer and same-layer context fusion. Experiments on NEU-DET and a self-built production-line backlight panel defect dataset show that HEC-YOLO reduces the parameters and FLOPs of YOLOv8n by 42.9% and 50.6%, respectively, while improving mAP@0.5 by 2.1 and 2.8 percentage points on the two datasets. On Jetson AGX Orin, it reaches 70.1 FPS while reducing per-image energy consumption and GPU temperature rise by 30.95% and 21.87%, respectively, relative to YOLOv8n.