Neural implicit functions including signed distance functions (SDFs) and unsigned distance functions (UDFs) have shown powerful ability in fitting the shape geometry. However, inferring continuous distance fields from discrete unoriented point clouds still remains a challenge. The neural network typically fits the shape with a rough surface and omits fine-grained geometric details such as shape edges and corners. In this paper, we propose a novel non-linear implicit filter to smooth the implicit field while preserving high-frequency geometry details. Our novelty lies in that we can filter the surface (zero level set) by the neighbor input points with gradients of the signed distance field. By moving the input raw point clouds along the gradient, our proposed implicit filtering can be extended to non-zero level sets to keep the promise consistency between different level sets, which consequently results in a better regularization of the zero level set. Since the unsigned distance function is non-differentiable at the zero level set and lacks a stable gradient field, we further propose a gradient immutable training schema to migrate the filter to the unsigned distance function learned from point clouds. By leveraging the UDF training schema, we also improve sparse-view reconstruction results. We conduct comprehensive experiments in surface reconstruction from objects, complex scene point clouds, and multi-view images, and we further extend to the point normal estimation and point cloud upsampling tasks. The numerical and visual comparisons demonstrate our improvements over the state-of-the-art methods under the widely used benchmarks.

While it is relatively easier to train humanoid robots to mimic specific locomotion skills, it is more challenging to learn from various motions and adhere to continuously changing commands. These robots must accurately track motion instructions, seamlessly transition between a variety of movements, and master intermediate motions not present in their reference data. In this work, we propose a novel approach that integrates human-like motion transfer with precise velocity tracking by a series of improvements to classical imitation learning. To enhance generalization, we employ the Wasserstein divergence criterion (WGAN-div). Furthermore, a Hybrid Internal Model provides structured estimates of hidden states and velocity to enhance mobile stability and environment adaptability, while a curiosity bonus fosters exploration. Our comprehensive method promises highly human-like locomotion that adapts to varying velocity requirements, direct generalization to unseen motions and multitasking, as well as zero-shot transfer to the simulator and the real world across different terrains. These advancements are validated through simulations across various robot models and extensive real-world experiments.

In this article, an adaptive fixed-time event-triggered consensus tracking control strategy is proposed for the robotic multiagent systems (MASs). First, this article considers the robotic MASs rather than the single robotic manipulator system, which is of great research significance in practical applications. Then, the adaptive fixed-time control method within the backstepping technique is developed such that each robotic manipulator can track the ideal signal more quickly. Moreover, in the face of complex tasks, the communication resources of the robotic MASs are in short supply. By sampling the data from the original controller, the relative threshold event-triggered control (RTETC) strategy is adopted for each robotic manipulator system, which can ensure that all signals in the closed-loop system are bounded without the Zeno phenomenon. In the end, a simulation example is presented to demonstrate the validity of the proposed control strategy.

Implicit neural networks have emerged as a crucial technology in 3D surface reconstruction. To reconstruct continuous surfaces from discrete point clouds, encoding the input points into regular grid features (plane or volume) has been commonly employed in existing approaches. However, these methods typically use the grid as an index for uniformly scattering point features. Compared with the irregular point features, the regular grid features may sacrifice some reconstruction details but improve efficiency. To take full advantage of these two types of features, we introduce a novel and high-efficiency attention mechanism between the grid and point features named Point-Grid Transformer (GridFormer). This mechanism treats the grid as a transfer point connecting the space and point cloud. Our method maximizes the spatial expressiveness of grid features and maintains computational efficiency. Furthermore, optimizing predictions over the entire space could potentially result in blurred boundaries. To address this issue, we further propose a boundary optimization strategy incorporating margin binary cross-entropy loss and boundary sampling. This approach enables us to achieve a more precise representation of the object structure. Our experiments validate that our method is effective and outperforms the state-of-the-art approaches under widely used benchmarks by producing more precise geometry reconstructions. The code is available at https://github.com/list17/GridFormer.

This article solves the cooperative adaptive tracking control problem for nonlinear pure-feedback multi-agent systems (MASs). Compared with the previous achievements of adaptive control of pure-feedback MASs, the partial derivative of the nonaffine function may not exist by using decoupling technology. In the controller design framework based on the backstepping technique, the additional state variables are processed using the special properties of the radial basis function neural networks (RBF NNs). A special-shaped Laplacian matrix is proposed to unify the leader gain form in the tracking error design process (the coefficient in the second term of tracking error). Furthermore, an event trigger mechanism (ETM) is introduced to save resources. The constructed controller under the ETM can not only stabilize the system states but also make the tracking error reach a small accuracy. Finally, the simulation results demonstrated the feasibility of the proposed method.

With the rapid advance of Internet of Things (IoT), it is difficult for cloud-centric computing to meet the requirements of low latency and ease of use. As an open and distributed system, edge computing integrates computing, networking, storage, and applications. It provides intelligent services on the edge of an IoT. The edge network is composed of various wireless and wired networks, and the computing and storage resources of edge nodes are limited. These conditions make the edge network expose to a variety of cyber attacks. Additionally, it is difficult for an IoT edge node to support large-scale network data collection and detection for IoT security. Although big data-enabled intrusion detection algorithms can ensure the high accuracy of intrusion detection systems, it is stressful for resource-limited edge nodes to implement those algorithms in IoT. Motivated by these challenges, we propose an intelligent intrusion detection algorithm implemented by big data mining based on a fuzzy rough set, generative adversarial network (GAN), and convolutional neural network (CNN). In our method, we first propose a fuzzy rough set-based algorithm to perform feature selection for big data via IoT. Then, we take advantage of the efficient feature extraction capabilities of CNN for implementing intrusion detection based on selected features. Furthermore, after combining CNN and GAN, we propose an intelligent algorithm to realize intrusion detection in a variety of scenarios. Finally, the proposed method is compared with existing methods for evaluation. Simulation results show that our method has up to 4% higher accuracy than existing methods.

The Social Internet of Things (SIoT) now penetrates our daily lives. As a strategy to alleviate the escalation of resource congestion, collaborative edge computing (CEC) has become a new paradigm for solving the needs of the Internet of Things (IoT). CEC can provide computing, storage, and network connection resources for remote devices. Because the edge network is closer to the connected devices, it involves a large amount of users’ privacy. This also makes edge networks face more and more security issues, such as Denial-of-Service (DoS) attacks, unauthorized access, packet sniffing, and man-in-the-middle attacks. To combat these issues and enhance the security of edge networks, we propose a deep learning-based intrusion detection algorithm. Based on the generative adversarial network (GAN), we designed a powerful intrusion detection method. Our intrusion detection method includes three phases. First, we use the feature selection module to process the collaborative edge network traffic. Second, a deep learning architecture based on GAN is designed for intrusion detection aiming at a single attack. Finally, we propose a new intrusion detection model by combining several intrusion detection models that aim at a single attack. Intrusion detection aiming at multiple attacks is realized through the designed GAN-based deep learning architecture. Besides, we provide a comprehensive evaluation to verify the effectiveness of the proposed method.

Intelligent Internet of Things (IIoT) is comprised of various wireless and wired networks for industrial applications, which makes it complex and heterogeneous.The openness of IIoT has led to the intractable problems of network security and management. Many network security and management functions rely on network traffic prediction techniques, such as anomaly detection and predictive network planning. Predicting IIoT network traffic is significantly difficult because its frequently updated topology and diversified services lead to irregular network traffic fluctuations. Motivated by these observations, we proposed a reinforcement learning-based mechanism in this article. We modeled the network traffic prediction problem as a Markov decision process, and then, predicted network traffic by Monte Carlo Q-learning. Furthermore, we addressed the real-time requirement of the proposed mechanism and we proposed a residual-based dictionary learning algorithm to improve the complexity of Monte Carlo Q-learning. Finally, the effectiveness of our mechanism was evaluated using the real network traffic.

Industrial Internet of Things (IIoT), as a common industrial application of Internet of Things, has been widely deployed in recent years. End-to-end network traffic is an essential information for many network security and management functions. This article investigates the issues of IIoT-oriented backbone network traffic prediction. Predicting the traffic of IIoT backbone networks is intractable because of the large number of prior network traffic information, which needs to consume expensive network resources for sampling. Motivated by that, we propose an effective prediction mechanism using multitask learning (MTL), which is a special paradigm of transfer learning. A deep learning architecture constructed by MTL and long short-term memory is designed. This deep architecture takes advantage of link loads as additional information to improve prediction accuracy. We provide a theoretical analysis for the MTL mechanism. The effectiveness is evaluated by implementing our mechanism on real network.