Vision-Language Models (VLMs) have great potential for advancing echocardiography (Echo) probe navigation, which is crucial for assisting sonographers in standardized view acquisition. However, the high clinical deployment costs and spurious correlations pose major challenges for VLM-based probe navigation. To address these challenges, we propose Contrastive Language-Image Navigator for Latent Hyperbolic Geometric Knowledge Distillation (CLINav-GKD), a novel VLM-based 6-DOF Echo probe navigation framework. Specifically, Contrastive Language-Image Navigator (CLN) proposes a lightweight VLM-based 6-DOF navigator, reducing deployment costs while improving sensitivity to quality variations. Latent Hy perbolic Geometric Distiller (LGD) models the global geometric-topology between samples, mitigating spurious correlations and enhancing robustness. We train CLINav-GKD on real-world data with probe motion trajectories. Experimental results show that CLINav-GKD outperforms other VLM-based distillation methods by 2.8%, 3.8%, and 3.6% in probe navigation, achieving a superior balance of accuracy, robustness, and deployability for real-world clinical use. Code and data are available at https://github.com/DaisyLi0516/CLINav-GKD.

Semantic communication (SemCom) shifts the focus from bit-level accuracy to the preservation of meaning, enabling more efficient and robust transmission. To achieve a high utilization of the wireless channel in SemCom, in this paper, we propose a channel assignment approach for polar code-based SemCom that allocates polarized channels according to semantic importance. Specifically, by combining eye-tracking data with semantic segmentation, we define two metrics that capture the contribution and correlation of semantic entities within an image. Leveraging these semantic metrics and polarized channel reliabilities, we formulate a constrained 0-1 optimization problem for polarized channel assignment and develop a priority-based algorithm that dynamically prioritizes semantically important content. Simulation results demonstrate that our method significantly outperforms the traditional channel allocation policy, especially under harsh channel conditions, by preserving critical visual information while reducing overall transmission redundancy.

The Entrance Dependent Vehicle Routing Problem (EDVRP) is a variant of the Vehicle Routing Problem (VRP) where the scale of cities influences routing outcomes, necessitating consideration of their entrances. This paper addresses EDVRP in agriculture, focusing on multi-parameter vehicle planning for irregularly shaped fields. To address the limitations of traditional methods, such as heuristic approaches, which often overlook field geometry and entrance constraints, we propose a Joint Probability Distribution Sampling Neural Network (JPDS-NN) to effectively solve the EDVRP. The network uses an encoder-decoder architecture with graph transformers and attention mechanisms to model routing as a Markov Decision Process, and is trained via reinforcement learning for efficient and rapid end-to-end planning. Experimental results indicate that JPDS-NN reduces travel distances by 48.4–65.4%, lowers fuel consumption by 14.0–17.6%, and computes two orders of magnitude faster than baseline methods, while demonstrating 15–25% superior performance in dynamic arrangement scenarios. Ablation studies validate the necessity of cross-attention and pre-training. The framework enables scalable, intelligent routing for large-scale farming under dynamic constraints.

Blockchain is an eminent technique to enhance the safety and robustness of the Internet of Things (IoT) network, due to its traits of decentralisation and transparency. Practical Byzantine Fault Tolerance (PBFT) blockchain consensus mechanism is well suited for wireless networks because of its low-computing requirement, low latency and high throughput. In this paper, we investigate the implementation of the base station (BS)-enabled wireless PBFT network, where the inter-node communications go through the BS in the normal case operation. The performance under such a scheme is analysed and evaluated through three metrics: consensus success probability, communication complexity, and average node transmit power. Results show that the proposed framework achieves higher scalability, lower communication complexity, and lower average node transmit power.

Predicting the future trajectories of other agents in the scene fast and effectively is crucial for autonomous driving systems. We note that high-quality predictions require us to take into account the subjective initiative of the target agents, which is reflected by the fact that they themselves make decisions based on their own predictions about the future, just like our ego vehicle's prediction-planning system. However, this characteristic has been neglected in previous studies. We introduce Look Before You Drive (LBYD), a two-stage approach that explicitly incorporates both past observations and future estimates to make predictions. To get a preliminary estimate of the future, we propose a neat and effective baseline capable of making predictions for multiple agents simultaneously. We use only the most basic structures, mainly Transformer, to ensure sufficient inference speed and room for expansion. On this basis, we cooperatively train two networks to enable the coarse estimates to boost final forecasting. Our experiments demonstrate that LBYD can significantly surpass the baseline performance. Moreover, while state-of-the-art methods rely on considering heterogeneity and artificially designed inductive biases for attention modeling, LBYD performs on par with SOTA without them on both the Argoverse 1 and the large scale Argoverse 2 datasets, and can run at 67 FPS on an RTX 3090 GPU.

Robotic grasping faces new challenges in human-robot-interaction scenarios. We consider the task that the robot grasps a target object designated by human's language directives. The robot not only needs to locate a target based on vision-and-language information, but also needs to predict the reasonable grasp pose candidate at various views and postures. In this work, we propose a novel interactive grasp policy, named Visual-Lingual-Grasp (VL-Grasp), to grasp the target specified by human language. First, we build a new challenging visual grounding dataset to provide functional training data for robotic interactive perception in indoor environments. Second, we propose a 6- Dof interactive grasp policy combined with visual grounding and 6- Dof grasp pose detection to extend the universality of interactive grasping. Third, we design a grasp pose filter module to enhance the performance of the policy. Experiments demonstrate the effectiveness and extendibility of the VL-Grasp in real world. The VL-Grasp achieves a success rate of 72.5 % in different indoor scenes. The code and dataset is available at https://github.com/luyh20/VL-Grasp.

The centralized system becomes less efficient, secure, and resilient as the network size and heterogeneity increase due to its inherent single point of failure issues. Distributed consensus mechanisms characterized by decentralization, autonomy, parallelism, and fault-tolerance can meet the increasing demands of safety and security in critical interconnected systems. This article establishes a Node and Link probabilistic failure model in the presence of node and communication link failures for a representative crash fault-tolerant distributed consensus protocol: RAFT. The analytical results in terms of the probability density function and the mean value of consensus reliability are derived. Two important reliability performance indicators, Reliability Gain and Tolerance Gain are proposed to indicate the linear relationship between the consensus reliability and two basic parameters, i.e., the joint failure rate and the maximum number of tolerant faulty nodes, which provide the theoretical guidance for quickly deploying an RAFT system. The special case of a distributed consensus network with already a certain number of failures and its adverse impact are evaluated. The Markov probabilistic models, definitions of Reliability Gain and Tolerance Gain, and the analysis methods proposed in this article can be extended to other consensus mechanisms.

Connected critical autonomous systems (C-CASs) are envisioned to significantly change our life and work styles through emerging vertical applications, such as autonomous vehicles and cooperative robots. However, as the scale of the connected nodes continues to grow, their heterogeneity and cybersecurity threats are more eminent, and conventional centralized communications and decision-making methodology are reaching their limit. This article is the first exploration of a trustworthy and fault-tolerant framework for C-CAS for achieving hyperreliable global decision making in a trustless environment, where the connected sensors/nodes are less reliable due to either communication failure or local decision error (e.g., by sensing algorithm/AI, etc.). The proposed framework is based on two iconic distributed consensus (DC) mechanisms: 1) practical Byzantine fault tolerance (PBFT) and 2) Raft, under the proposed perception-initiative-consensus-action (PICA) protocol with wireless connections among the nodes. We first analytically derived consensus reliability in six different system models. The other fundamental performance metrics, such as the consensus throughput and latency, node scalability, and reliability gain are also analytically derived. These analytical results provide basic design guidelines for wireless DC (WDC) usage in the C-CAS systems. The results show that WDC significantly improves overall system reliability with the increasing number of participating nodes.

In this paper, the Capacitated Vehicle Routing Problem (CVRP) of multi-depot express delivery is investigated based on the actual express delivery business in Beijing and driving intention-based road network. An Adaptive Simulated Annealing and Artificial Fish Swarm Algorithm (A-SAAFSA) is proposed to solve the CVRP. The basic ideas are use a “certainty” probability to accept the worst solution through the Metropolis criterion in the search process, and a strategy of adjusting the swimming direction to avoid falling into the local optimal solution. Moreover, an adaptive visual strategy, which adjusts the visual range adaptively in real time according to the current solution quality, is used to ensure the efficient searching and accuracy of the algorithm. Experimental results show that the A-SAAFSA algorithm outperforms four well-known algorithms, namely simulated annealing and artificial fish swarm algorithm, artificial fish swarm algorithm, simulated annealing algorithm, and genetic algorithm.