Edge intelligence, the fusion of edge computing and artificial intelligence (AI), drives the advancement of intelligent Internet of Things (IoT). Since AI applications are often data-and computation-intensive, resource-scarce edge devices need to migrate data to resource-rich edge servers through computation offloading to meet requirements such as energy efficiency and low latency. Existing studies often focus on CPU-based edge systems and neglect the impacts of other components, such as memory, on offloading. From a parallel processing perspective, this article establishes a system model and a multi-objective optimization model for edge intelligence systems with heterogeneous components, including diverse processors, memory, network, and applications, to minimize system energy consumption, total execution time, and the workload ratio of edge servers. A multi-objective optimization algorithm integrating archive initialization, hybrid perturbation, clustering, and modified simulated annealing is proposed and validated through experiments using real-world software and hardware. Results demonstrate that the proposed algorithm significantly outperforms comparative algorithms in terms of inverted generational distance, pure diversity, and run-time while revealing the influence of application characteristics on offloading performance.

Due to the diversity of edge devices (EDs) and applications, edge systems are heterogeneous and have been applied in artificial intelligence fields, such as smart factories and intelligent transportation, which is called heterogeneous edge intelligence. Many studies employ computation offloading to transfer processing data from resource-scarce EDs to resource-rich edge servers. These studies primarily focus on the overall resource consumption of homogeneous edge systems, neglecting the system heterogeneity and the details of resource consumption. In this article, we construct a system model from a parallel perspective for the heterogeneous edge system with different processors, memory, and applications, which perceives the cost of energy and delay from three levels: system, application, and component. A hybrid metaheuristic algorithm combined with a greedy rule, hybrid mutation, and whale optimization algorithm (GHMWOA) is proposed to realize partial computation offloading. A partial offloading architecture of heterogeneous edge intelligence is proposed to validate our model and algorithm with real-world hardware and software. Experiment results not only show GHMWOA outperforms multiple classical optimization algorithms in minimizing energy consumption, but also discover on which system component energy consumption depends, and how properties of application and system influence the cost of energy.

Multidimensional Cockpit Perception (MCP) is an indispensable technology in assisted driving systems for modern vehicles, which requires joint recognition of the driver's emotional behaviors, traffic context, and vehicle condition. Despite promising progress in 2D to 3D neural networks, semantic discrepancies and intricate temporal dependencies among different signal streams still cause serious performance bottlenecks. To address these challenges, this paper proposes a Mutual Information-guided signal Graph fusion (MIG) framework for robust cockpit perception. MIG introduces an adaptive mutual information constraint to maximize the mutual information among different perceived signal streams, thus reinforcing the task-related feature semantic consistency. In addition, a signal graph fusion module is designed to capture intra- and inter-signal elemental correlations at a fine-grained level to learn joint multi-signal representations used for different downstream tasks. Systematic experiments on the MCP benchmark verify the effectiveness of the proposed framework and the necessity of our components.

Surface electromyography (sEMG) signals-based gesture recognition method is widely employed in human-computer interaction task. In this paper, we proposed a phase locked value (PLV)-based feature extraction method for sEMG-based gesture recognition. We conducted validation experiments on public datasets by integrating the previous proposed gesture recognition algorithms, STCN-GR and ConSSL. PLV-STCN-GR and PLV-ConSSL were utilized to test the performance of proposed feature extraction method in intra-session, inter-session and inter-subject scenarios. In general, the proposed PLV-based sEMG decoding framework has achieved good recognition results on public datasets and demonstrated the possibility of its application in practical scenarios.

Due to the advantages of high flexibility, large workspace, and good human-body compatibility, flexible tendon-driven surgical continuum robots have attracted a lot of attention in robot-assisted minimally invasive surgery. However, due to the coupling of the position and angle of the continuum robot, and the easy deformation of the external force, its inverse kinematics solution has always been a challenge. This paper proposes a fast inverse kinematics solver for surgical continuum robots with a variable curvature model. Firstly, the deformation of the continuum robot is analyzed, and a representation method of the variable curvature model is proposed. Next, to solve the inverse kinematics problem when the continuum robot deforms under load, FABRIKv is proposed by improving the Forward And Backward Reaching Inverse Kinematics (FABRIK). During the inverse kinematics solution, the algorithm preserves the real-time nature of FABRIK and corrects for deformation effects caused by the load. Finally, the experiment verifies the rationality and effectiveness of the variable curvature model representation method, as well as the fastness and accuracy of the FARIKv solver.

The ability to perform efficient trajectory planning is crucial for UAV to carry out tasks autonomously. However, existing research on UAV trajectory planning often employs the cascade process method that involves high-precision maps, real-time positioning and path planning. These methods have limitations such as high computational complexity and time delay, which hinder the efficiency of trajectory planning. End-to-end trajectory planning methods offer a promising solution to this problem. As the core of these end-to-end methods, perception-end plays a decisive role in trajectory planning. But current multimodal fusion of perception is only post-fusion, lacks intermediate feature-level fusion and lacks attention to global visuospatial information. To solve these problems, we propose a new network architecture called FlyTransformer, which fuses the proprioceptive state and visual perception in feature-level for end-to-end trajectory planning. And the key visuospatial information can be attentioned in this architecture. We evaluate our method in forest and cuboid scenarios and their corresponding outdoor scenarios. The results show that FlyTransformer outperforms other baseline algorithms in terms of efficiency and performance.

Gesture recognition using surface electromyography (sEMG) shows its great potential in the field of human-computer interaction (HCI). Previous works achieve relatively good performance based on the assumption of invariant statistic distribution. However, the practical application effect is unsatisfactory due to the problem of domain shift. Existing approaches need plenty of labeled sEMG samples from target scenarios for calibration, which is burdensome for experimenters and users. In this work, we present a contrastive self-supervised learning framework (ConSSL) for sEMG-based gesture recognition to realize domain adaptation in target domains. After pretraining on a bunch of unlabeled samples, only a small number of labeled samples are needed for calibration and domain adaptation. Experimental results indicate that the proposed framework out-performs other approaches even$if\leq 50\%$labeled samples in target scenarios are available and achieves the state-of-the-art.

Reinforcement learning with adversarial training is currently a key method for improving the robustness of DRL. However, in adversarial training, especially for unstable or disturbance-sensitive systems, the adversary always learns the policy significantly faster than the DRL agent and thus easily generates powerful perturbations. The agent cannot effectively adapt to the overly powerful adversary, which leads to unstable training and even failure to learn the robust policy. In this work, we propose a novel adversarial training method, called Curriculum Adversarial Training, inspired by the idea of curriculum learning. The method dynamically adjusts the strength of the adversary through natural curriculum learning for progressive adversarial training. Thus, the DRL system considers to reasonable learning rules, and the agent faces a suitable learning process. Furthermore, we adopt an advanced action space perturbation method with a most attractive ability as the adversary during training. The proposed method is compared with popular baseline methods through MuJoCo tasks. Experimental results show that our method can improve the robustness of the policy significantly and adapt to uncertain environment effectively.

The paper presents an object detection accelerator featuring a processing-in-memory (PIM) architecture on FPGAs. PIM architectures are well known for their energy efficiency and avoidance of the memory wall. In the accelerator, a PIM unit is developed using BRAM and LUT based counters, which also helps to improve the DSP performance density. The overall architecture consists of 64 PIM units and three memory buffers to store inter-layer results. A shrunk and quantized Tiny-YOLO network is mapped to the PIM accelerator, where DRAM access is fully eliminated during inference. The design achieves a throughput of 201.6 GOPs at 100MHz clock rate and correspondingly, a performance density of 0.57 GOPS/DSP.

This paper presents a design strategy of chiplet-based processing-in-memory systems for deep neural network applications. Monolithic silicon chips are area and power limited, failing to catch the recent rapid growth of deep learning algorithms. The paper first demonstrates a straightforward layer-wise method that partitions the workload of a monolithic accelerator to a multi-chiplet pipeline. A quantitative analysis shows that the straightforward separation degrades the overall utilization of computing resources due to the reduced on-chiplet memory size, thus introducing a higher memory wall. A tile interleaving strategy is proposed to overcome such degradation. This strategy can segment one layer to different chiplets which maximizes the computing utilization. To facilitate the strategy, the modification of the chiplet system hardware is also discussed. To validate the proposed strategy, a nine-chiplet processing-in-memory system is evaluated with a custom-designed object detection network. Each chiplet can achieve a peak performance of 204.8GOPS at a 100-MHz rate. The peak performance of the overall system is 1.711TOPS, where no off-chip memory access is needed. By the tile interleaving strategy, the utilization is improved from 53.9 to 92.8