Jianwu Dang

Augmented reality (AR) enhances the real world scene by overlaying virtual information onto it. Vision-based 3D tracking and registration is the key technology for ensuring the fusion of virtual and real content in monocular AR systems. Existing mainstream visual tracking and registration methods are susceptible to illumination variations, motion blur, target occlusion, and dynamic background interference in complex scenarios. They also suffer from low computational efficiency, cumulative pose errors, and insufficient stability, making them difficult to deploy on low power edge devices such as embedded systems and mobile terminals. To address these issues, this paper proposes a lightweight monocular AR 3D tracking and registration method that integrates ORB-FREAK features, mismatching outlier filtering, background weighted mean shift, and template-based relocalization. The method does not rely on depth sensors or neural network inference, enabling efficient and accurate lightweight pose estimation. Specifically, we first combine the ORB (Oriented FAST and Rotated BRIEF) descriptor with the FREAK (Fast Retina Keypoint) algorithm for feature detection and initial matching. Hamming distance is used for coarse filtering of mismatched point pairs, and an ascending sort combined with an iterative sequential sampling strategy is applied to solve the optimal homography matrix, significantly improving the accuracy and efficiency of matrix estimation. Then, distance constraints among feature points are imposed on the target registration region to optimize the selection, and camera pose is computed based on the matching between 2D feature points and their corresponding 3D spatial coordinates, eliminating the error accumulation problem of conventional algorithms. Real-time feature matching is further used to correct the optical flow tracking sequence and camera pose, ensuring the continuity of the AR tracking process. Finally, a background weighted mean shift algorithm is introduced to narrow the feature detection range and suppress background interference, complemented by a template-matching relocalization module and a dynamic model update strategy, which effectively enhance the robustness of continuous tracking and registration under complex conditions. Experimental results demonstrate that, in extreme scenarios such as low light conditions, high speed motion, and occlusion, the proposed method achieves AR 3D tracking and registration success rates of 86.7%, 82.3%, and 78.5%, respectively. It exhibits superior performance in pose estimation accuracy and anti-interference capability in complex environments, with significantly reduced computational overhead. Moreover, it can achieve robust and continuous AR 3D tracking and registration on low power edge devices, effectively adapting to demanding AR application scenarios and providing reliable technical support for lightweight AR applications.

Gesture interaction allows users to interact with objects through natural hand movements without direct physical contact. However, in gesture interaction, the hand moves frequently with large variations in direction, and hand joints are prone to self-occlusion. The accuracy of existing 3D point cloud gesture estimation methods is insufficient to meet the requirements of natural gesture interaction. This article proposes a 3D point cloud gesture estimation method based on EdgeConv to reconstruct joint features. The method first converts depth maps into point cloud data to reduce the impact of viewpoint variations on depth map representation of the same gesture, thereby mitigating the effect on estimation accuracy. Then, the global features are reconstructed using EdgeConv in the initial joint estimation module so that the reconstructed features contain structural information between hand joints. Finally, local joint refinement is performed by constructing local features centered around the refined joint twice using EdgeConv. These features include reference information from within group points to group center points and structural information between joint points. Using EdgeConv to enrich hand joint feature information can effectively improve the performance of hand joint estimation. Comparative experimental analysis was conducted on the 3D point cloud datasets ICVL, NYU, and MSRA, and complex scene experimental analysis was conducted on the InterHand2.6M dataset. Furthermore, the development of virtual–real interaction applications was carried out. The experimental results show that the 3D point cloud gesture estimation method proposed in this paper has high accuracy, strong generalization and robustness, providing solid support for natural gesture interaction applications.

Reconstructing hand and object shapes from a single view during interaction remains challenging due to severe mutual occlusion and the need for high physical plausibility. To address this, we propose a novel framework for hand–object interaction reconstruction based on holistic, multi-stage collaborative optimization. Unlike methods that process hands and objects independently or apply constraints as late-stage post-processing, our model progressively enforces physical consistency and geometric accuracy throughout the entire reconstruction pipeline. Our network takes an RGB-D image as input. An adaptive feature fusion module first combines color and depth information to improve robustness against sensing uncertainties. We then introduce structural priors for 2D pose estimation and leverage texture cues to refine depth-based 3D pose initialization. Central to our approach is the iterative application of a dense mutual attention mechanism during sparse-to-dense mesh recovery, which dynamically captures interaction dependencies while refining geometry. Finally, we use a Signed Distance Function (SDF) representation explicitly designed for contact surfaces to prevent interpenetration and ensure physically plausible results. Through comprehensive experiments, our method demonstrates significant improvements on the challenging ObMan and DexYCB benchmarks, outperforming state-of-the-art techniques. Specifically, on the ObMan dataset, our approach achieves hand CD_h and object CD_o metrics of 0.077 cm² and 0.483 cm², respectively. Similarly, on the DexYCB dataset, it attains hand CD_h and object CD_o values of 0.251 cm² and 1.127 cm², respectively.

In recent years, telecom fraud remains prevalent in many regions, severely impacting people’s daily lives and causing substantial economic losses. However, previous research has mainly relied on expert knowledge for feature engineering, which lags behind and struggles to adapt to the continuously evolving patterns of fraud effectively. In addition, the extreme imbalance in fraud amounts within real communication data hinders the development of deep learning methods. In response, we propose a feature transformation method to represent users’ communication behavior as comprehensively as possible, and develop a convolutional neural network (CNN) with a Focal Loss function to identify rare fraudulent activities in highly imbalanced data. Experimental results on a real-world dataset show that, under conditions of severe class imbalance, the proposed method significantly outperforms existing approaches in two key metrics: recall (0.7850) and AUC (0.8662). Our work provides a new approach for telecommunication fraud detection, enabling the effective identification of fraudulent numbers.

The rail transit switch machine ensures the safe turning and operation of trains on the track by switching switch positions, locking switch rails, and reflecting switch status in real time. However, in the detection of complex rail transit switch machine parts such as augmented reality and automatic inspection, existing algorithms have problems such as insufficient feature extraction, large computational complexity, and high demand for hardware resources. This article proposes a complex scene rail transit switch machine parts detection network YOLO-SMPDNet (YOLO-based Switch Machine Parts Detecting Network). The YOLOv8s backbone network is improved, and the number of network parameters are reduced by introducing MobileNetV3. Then a parameter-free attention-enhanced ResAM module is designed, which forms a lightweight detection network with the improved network, improving detection efficiency. Finally, Focal IoU Loss is introduced to more accurately define the scale information of the prediction box, alleviate the problem of imbalanced positive and negative samples, and improve the relative ambiguity of CIoU Loss in YOLOv8s on the definition of aspect ratio. By validating the performance of YOLO-SMPDNet on a self-made dataset of rail transit switch machines, the results show that YOLO-SMPDNet can significantly improve detection accuracy and real-time performance and has robust comprehensive detection capabilities for rail transit switch machine parts and good practical application performance.

The existing learning resource recommendation systems suffer from data sparsity and missing data labels, leading to the insufficient mining of the correlation between users and courses. To address these issues, we propose a learning resource recommendation method based on graph contrastive learning, which uses graph contrastive learning to construct an auxiliary recommendation task combined with a main recommendation task, achieving the joint recommendation of learning resources. Firstly, the interaction bipartite graph between the user and the course is input into a lightweight graph convolutional network, and the embedded representation of each node in the graph is obtained after compilation. Then, for the input user–course interaction bipartite graph, noise vectors are randomly added to each node in the embedding space to perturb the embedding of graph encoder node, forming a perturbation embedding representation of the node to enhance the data. Subsequently, the graph contrastive learning method is used to construct auxiliary recommendation tasks. Finally, the main task of recommendation supervision and the constructed auxiliary task of graph contrastive learning are jointly learned to alleviate data sparsity. The experimental results show that the proposed method in this paper has improved the Recall@5 by 5.7% and 11.2% and the NDCG@5 by 0.1% and 6.4%, respectively, on the MOOCCube and Amazon-Book datasets compared with the node enhancement methods. Therefore, the proposed method can significantly improve the mining level of users and courses by using a graph comparison method in the auxiliary recommendation task and has better noise immunity and robustness.

Ensuring trust, security, and privacy among all participating parties in the process of sharing supervision data is crucial for engineering quality and safety. However, the current centralized architecture platforms that are commonly used for engineering supervision data have problems such as low data sharing and high centralization. A blockchain-based framework for the secure sharing of engineering supervision data is proposed by utilizing the tamper-proof, decentralized, and traceable characteristics of blockchain. The secure storage of supervision data is achieved by combining it with the IPFS (InterPlanetary File System), reducing the storage pressure of on-chain data. Additionally, a fast data retrieval framework is designed based on the storage characteristics of supervision data. Then, CP-ABE (Ciphertext Policy Attribute Based Encryption) is combined with a data storage framework to ensure the privacy, security, and reliability of supervisory data during the sharing process. Finally, smart contracts are designed under the designed framework to ensure the automatic and trustworthy execution of access control processes. The analysis and evaluation results of the security, encryption and decryption, and cost performance of the proposed blockchain framework show that the encryption and decryption time is completed within 0.1 s, the Gas cost is within the normal consumption range, and the time cost of smart contract invocation does not exceed 5 s, demonstrating good availability and reusability of the method proposed in this article.

Object 6D pose estimation, as a key technology in applications such as augmented reality (AR), virtual reality (VR), robotics, and autonomous driving, requires the prediction of the 3D position and 3D pose of objects robustly from complex scene images. However, complex environmental factors such as occlusion, noise, weak texture, and lighting changes may affect the accuracy and robustness of object 6D pose estimation. We propose a robust CoS-PVNet (complex scenarios pixel-wise voting network) pose estimation network for complex scenes. By adding a pixel-weight layer based on the PVNet network, more accurate pixel point vectors are selected, and dilated convolution and adaptive weighting strategies are used to capture local and global contextual information of the input feature map. At the same time, the perspective-n-point localization algorithm is used to accurately locate 2D key points to solve the pose of 6D objects, and then, the transformation relationship matrix of 6D pose projection is solved. The research results indicate that on the LineMod and Occlusion LineMod datasets, CoS-PVNet has high accuracy and can achieve stable and robust 6D pose estimation even in complex scenes.

Extended reality (XR) is a general term for virtual reality (VR), augmented reality (AR), and mixed reality (MR). By converting abstract digital expressions into intelligent feedback through figures, one can effectively compensate for the poor performance of traditional learning in deep cognitive processing and operational skills training. However, the extant results are uncertain, and only a limited number of studies have investigated the influence mechanism of heterogeneity among VR, AR, and MR on procedural knowledge learning, higher-level presence and generative cognitive processing, higher-quality operational behaviour and knowledge transfer. This study explored and analysed the influence of XR heterogeneity on procedural knowledge gain and operation training. Participants (n = 127) were recruited from among junior undergraduate students majoring in rail traffic signal and control. Based on the procedural disassembly and assembly of the ZD6 electric switch machine, an XR pre-training system and a 3D-printed combined virtual-real procedural training system were developed to collect three types of data regarding knowledge gain, procedural training and a set of questionnaires. Subsequently, a one-way analysis of variance was conducted with XR as the moderator variable. The results indicate that fully immersive VR is more conducive to procedural operation training; however, space and real factors must be considered. AR is consistent with the learner's operating habits, but the corresponding lack of immersion and interactivity restricts the improvement of procedural learning effects. MR multimodal perception leads to an increase in cognitive load, but moderate cognitive pressure is conducive to the cultivation of procedural cognition skills. Furthermore, this study proposes suggestions for using XR to intensify the perception and understanding of procedural cognitive actions.