Ye Wang, Qianglong Chen, Siyuan Wang, Zejun Li, Shijie Guo, Zhirui Zhang, Zhongyu Wei. Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2026.

Multi-task imitation learning is a pressing issue for robotics applications. However, the inaccuracy and inefficiency of imitation learning frequently limits current approaches. To cope with the problem, this paper proposes Modules Generation with Adaptive Discriminator-Feedback Mechanism with Diffusion Model (MGAD-FMDM), a novel general modular method for robotic multi-task imitation learning. A transferable adaptive discriminator constitutes the first component whose function is to automatically generate potential primitive modules, which enables robots to improve the imitation accuracy. A universal feedback mechanism for guiding policy updates represents the second development. By employing a diffusion model, the feedback mechanism achieves a two-step process of first recognizing the expert demonstrated distribution, then evaluating the fit between imitated trajectories and expert demonstrations, thereby enhancing learning efficiency. Extensive experiments on practical robot tasks indicate that the proposed MGAD-FMDM achieves superior performance over state-of-the-art methods, including the improvements of 2.22%-34.85% in Spearman’s rank correlation, 14.29%-100.00% in success rate, 5.43%-64.35% in earth mover’s distance, and 21.74%-30.10% in convergence speed. The video can be obtained at: https://github.com/yueli0827/Robotic-Multi-Task-Imitation-Learning.

The growing population of bedridden patients due to global aging has sharply increased the demand for adaptive head support systems. However, significant challenges, especially insufficient integration of perception and control, as well as limited fine-grained adjustment capabilities, remain in this domain. To overcome these challenges, a dual-model soft pneumatic control framework is proposed in this study. The proposed framework integrates a Genetic Algorithm-optimized Multilayer Perceptron (GA-MLP) for real-time head load perception and an Extreme Gradient Boosting (XGBoost) model for precise pressure adjustment. Algorithm fusion enables seamless coordination between dynamic sensing and intelligent actuation through the airbag platform. Furthermore, an Internet of Things (IoT)-enabled hardware system, which integrates soft pneumatic actuation and wireless communication, is developed to support remote monitoring and real-time control. Experimental results show high regression performance, with the coefficient of determination (R²) of 0.9881 and the mean absolute error (MAE) of 0.5321 for load estimation, and the R² of 0.9893 with the MAE of 0.4891 for angle adjustment. This study presents a novel soft pneumatic solution for load-perception-driven head posture adjustment in intelligent rehabilitation.

Recent advancements in artificial intelligence have enhanced the automation of redundant robots, extending their applications from safety cages to unstructured environments with diverse constraints. In such scenarios, these robots typically operate through sequential stages of motion planning and trajectory tracking. However, the decoupling of control objectives between these stages often results in cumulative error. To address this issue, a redundancy adaptive fuzzy controller is proposed. It leverages kinematic redundancy by utilizing redundant DOF to compensate for unknown dynamics through a Fuzzy Logic System (FLS), and achieving fixed-time convergence for task-space pose tracking. To generate constraint satisfying motion during the tracking task, virtual forces are designed in joint space, and are compensated in task space through FLS. Consequently, virtual forces can be integrated with the fuzzy controller to construct Motion Planning Controller (MoPC) that guarantees constraint satisfaction and accurate tracking motion. Additionally, a saturation system is introduced to define a task hierarchy for handling the situation where tracking objective is incompatible with constraints requirements. Lyapunov-based analysis is further conducted to guarantee system stability. The effectiveness of MoPC is verified by precise task-space trajectory tracking and constraint satisfaction of 7-DOF redundant robots in unstructured environments, using both simulations and physical experiments, demonstrating improved convergence rates, tracking accuracy, chattering reduction, and robustness.

Major depressive disorder (MDD) is a common neuropsychiatric disorder, yet its underlying physiological mechanisms remain unclear, limiting diagnostic advances. Functional connectivity (FC) derived from resting-state functional magnetic resonance imaging (rs-fMRI), when combined with deep learning methods, has shown promise as diagnostic biomarker. Currently, most FC-based diagnostic methods rely on graph structures modeled by FC, and are limited to pairwise interactions between brain regions. Hypergraph representations enable the characterization of higher-order interactions across multiple regions. However, existing hypergraph models ignore the directionality of these interactions, thus limiting their ability to capture complex neural dynamics. To address these limitations, this study proposes a dynamic weight aware directed hypergraph learning method - dwDHGL, for MDD identification and subtype analysis. dwDHGL captures asymmetric causal interactions by modeling temporal lag effects and constructs a directed hypergraph network(DHN). It further utilizes a self-attention mechanism to dynamically learn inter node interactions during message passing and adaptively differentiate node importance. A node weight aware directed hypergraph convolution is designed to aggregate features based on hyperedge directions, incorporating dynamic weights to enhance representation learning. The proposed method is evaluated on the large-scale REST-meta-MDD dataset, achieving an MDD identification accuracy of 73.75 %, and outperforming existing advanced methods in subtype identification. Furthermore, dwDHGL identifies discriminative directed hyperedges, with the inferior frontal gyrus triangular part emerging as key biomarkers, providing new insights into the neural mechanisms of MDD.

This paper explores the application of large language models (LLMs) in nursing and elderly care, focusing on AI-driven patient monitoring and interaction. We introduce a novel Chinese nursing dataset and implement incremental pre-training (IPT) and supervised fine-tuning (SFT) techniques to enhance LLM performance in specialized tasks. Using LangChain, we develop an interactable nursing assistant capable of real-time care and personalized interventions. Experimental results demonstrate significant improvements, paving the way for AI-driven solutions to meet the growing demands of healthcare in aging populations.

Unsupervised anomaly detection identifies deviations from normal patterns as anomalies. Recently, unsupervised methods have made significant strides in anomaly detection. However, single-scale feature extraction struggles to capture subtle anomalies and existing methods frequently emphasize exclusively on local information while disregarding global semantic information. In this paper, we propose a new normalizing flow called Multi-Level Normalizing Flow (MLFlow) for anomaly detection and localization. First, we input normal images and extract multi-scale features using a pre-trained feature extractor. Second, MLFlow receives the multi-scale feature maps and density estimate. StepFlow and ConvergeFlow are the two main modules of MLFlow. Specifically, the StepFlow independently transforms each layer of feature maps, allowing the lower layer to capture detailed features, the middle layer to extract local features and the top layer to extract semantic information. Additionally, the ConvergeFlow combines transformed multi-scale feature maps, enhancing the comprehensive analysis capability for anomalies. Experimental results on MVTec AD, BeanTech AD and VisA datasets reveal that the proposed method performs exceptionally well in anomaly detection and localization tasks, surpassing existing methods and achieving the state-of-the-art performance.

Imitation learning is a significant precondition for robotics automation. However, traditional approaches of Behavioral Cloning (BC), Inverse Reinforcement Learning (IRL) and Generative Adversarial Imitation Learning (GAIL) are frequently accompanied by the problems of poor learning success and the over-reliance on expert samples. To remedy these problems, this paper proposes a novel robotic trajectory imitation method, named TGOD-SD. First, the TGOD generates several distinguishable trajectories for imitation through noreward Reinforcement Learning. Then, the agent searches the most appropriate trajectory that matching the target trajectory by Sinkhorn Distance (SD). Finally, the proposed TGOD-SD is verified in real universal robot UR5. TGOD-SD can directly follow the expert demonstration, and also succeed the similar robotic tasks. Quantitative and qualitative evaluation illustrates that TGOD-SD achieves remarkable learning success rate compared with the state-of-the-art robot imitation learning methods. In addition, TGOD-SD achieves robot imitation learning from a single expert demonstration, effectively reducing the dependence on the expert demonstrations. The code and video can be obtained at: https://github.com/Nursing-Robot-Laboratory

Individual-specific calibration is a major bottleneck in motor imagery (MI) electroencephalogram (EEG) decoding, limiting real-world neural-feedback rehabilitation. Transfer learning, particularly Test-Time Adaptation (TTA), offers a promising solution for direct online cross-subject adaptation, handling sequentially arriving unlabeled MI-EEG data. However, existing TTA methods, primarily designed for domains such as computer vision, face challenges when applied to MI-EEG data due to its scarcity and non-stationary nature. To address the challenges in direct online MI-EEG decoding, this paper proposes MI-IASW, a novel framework combining Information-Aggregation (IA) and Source-Guided Pseudo-Label Weighting (SW). IA leverages Mixed and Adaptive Batch Normalization (MABN) to ensure effective aggregation of statistical and gradient information. Additionally, IA adopts a Weight Aggregation (WA) strategy to improve generalization under limited data. Meanwhile, SW first evaluates the overconfident pseudo-labels with the guidance of source centers and then employs Class-Aware Weighting (CAW) to adjust sample contributions to the loss function. Experimental evaluations on two public MI-EEG datasets demonstrate that our proposed framework outperforms various competitive baselines, achieving an average performance gain of 3.17% over the baseline TTA methods and 6.80% over the source model. By eliminating the need for individual-specific offline calibration, MI-IASW enables practical deployment in real-world rehabilitation and improves cross-subject decoding.

With the increasing complexity of pipeline systems in various industrial and environmental applications, there is a critical need for flexible and efficient robotic solutions that can navigate and inspect confined spaces. This paper introduces a lightweight pipeline crawling robot based on spring-roll dielectric elastomer actuators (DEAs). Inspired by the adaptability of caterpillars, the robot combines anisotropic friction feet with a spring-roll DEA structure to achieve high-speed movement. It operates effectively in pipes with diameters ranging from 16 mm to 20 mm, reaching a maximum speed of 357 mm/s (5.95 BL/s) under a 3.5 kV driving voltage. The optimized design enhances actuator performance and friction distribution, significantly outperforming existing soft crawling robots. This innovation demonstrates great potential for high-speed, lightweight pipeline inspection applications and advances the field of soft robotics for diverse industrial tasks.

BACKGROUND: With the global population aging and advancements in the medical system, long-term care in healthcare institutions and home settings has become essential for older adults with disabilities. However, the diverse and scattered care requirements of these individuals make developing effective long-term care plans heavily reliant on professional nursing staff, and even experienced caregivers may make mistakes or face confusion during the care plan development process. Consequently, there is a rigid demand for intelligent systems that can recommend comprehensive long-term care plans for older adults with disabilities who have stable clinical conditions. OBJECTIVE: This study aims to utilize deep learning methods to recommend comprehensive care plans for the older adults with disabilities. METHODS: We model the care data of older adults with disabilities using a bipartite graph. Additionally, we employ a prediction-based graph self-supervised learning (SSL) method to mine deep representations of graph nodes. Furthermore, we propose a novel graph Transformer architecture that incorporates eigenvector centrality to augment node features and uses graph structural information as references for the self-attention mechanism. Ultimately, we present the Bipartite Graph Transformer (BiT) model to provide personalized long-term care plan recommendation. RESULTS: We constructed a bipartite graph comprising of 1917 nodes and 195 240 edges derived from real-world care data. The proposed model demonstrates outstanding performance, achieving an overall F1 score of 0.905 for care plan recommendations. Each care service item reached an average F1 score of 0.897, indicating that the BiT model is capable of accurately selecting services and effectively balancing the trade-off between incorrect and missed selections. DISCUSSION: The BiT model proposed in this paper demonstrates strong potential for improving long-term care plan recommendations by leveraging bipartite graph modeling and graph SSL. This approach addresses the challenges of manual care planning, such as inefficiency, bias, and errors, by offering personalized and data-driven recommendations. While the model excels in common care items, its performance on rare or complex services could be enhanced with further refinement. These findings highlight the model's ability to provide scalable, AI-driven solutions to optimize care planning, though future research should explore its applicability across diverse healthcare settings and service types. CONCLUSIONS: Compared to previous research, the novel model proposed in this article effectively learns latent topology in bipartite graphs and achieves superior recommendation performance. Our study demonstrates the applicability of SSL and graph transformers in recommending long-term care plans for older adults with disabilities.

When humans perform pouring tasks, they exhibit consistent accuracy, regardless of the liquid type, container, or environmental conditions. This proficiency stems from their ability to effectively utilize both vision and hearing while also considering various factors. However, in the domain of robotic liquid pouring, the combination of multimodal information is effectively rarely leveraged to accomplish automatic control of robotic liquid pouring. To address this limitation, a multimodal information fusion network (MMFNet) is designed for estimating liquid height and pouring state. The MMFNet employs cross-attention networks and motion features to enhance visual features (VFs). Subsequently, multimodal transformers are utilized to fuse audio features with the enhanced VFs, enabling the MMFNet to estimate both liquid height and pouring state accurately. Finally, the detection results are combined with demonstration learning to make robots learn pouring motion trajectory encoded by the Gaussian mixture model (GMM). The experimental results demonstrate the effectiveness of MMFNet in significantly improving the detection accuracy of liquid height and pouring state. Furthermore, by employing the GMM enabled by MMFNet, robots can acquire robust pouring motion planning, enhancing their capabilities in performing pouring tasks.

In the context of computer vision and human-robot interaction, forecasting 3D human poses is crucial for understanding human behavior and enhancing the predictive capabilities of intelligent systems. While existing methods have made significant progress, they often focus on predicting major body joints, overlooking fine-grained gestures and their interaction with objects. Human hand movements, particularly during object interactions, play a pivotal role and provide more precise expressions of human poses. This work fills this gap and introduces a novel paradigm: forecasting 3D whole-body human poses with a focus on grasping objects. This task involves predicting activities across all joints in the body and hands, encompassing the complexities of internal heterogeneity and external interactivity. To tackle these challenges, we also propose a novel approach: C3HOST, cross-context cross-modal consolidation for 3D whole-body pose forecasting, effectively handles the complexities of internal heterogeneity and external interactivity. C3HOST involves distinct steps, including the heterogeneous content encoding and alignment, and cross-modal feature learning and interaction. These enable us to predict activities across all body and hand joints, ensuring high-precision whole-body human pose prediction, even during object grasping. Extensive experiments on two benchmarks demonstrate that our model significantly enhances the accuracy of whole-body human motion prediction. The project page is available at https://sites.google.com/view/c3host.

Iontronic tactile sensors are promising to measure spatial-temporal contact information with high performance. However, no suitable measuring method has been presented, due to issues with crosstalk and non-negligible equivalent resistance. Hence, this study presents an impedance-separating method, which does not require complex analog components. A general Quadri-Terminal Impedance Network (QTIN) model is introduced to reduce crosstalk, which has specific compatibility with the impedance-separating method. The precise ranges are measured, showing non-rectangle shapes suitable for the response of iontronic tactile sensors. A simple denoising method is provided to reduce initial array noise obviously. This work could benefit various scenarios, such as human-robot interaction and physiological information monitoring.

With the aging of society, there has been an increase in the number of elderly individuals with limb movement disorders. Active rehabilitation training using limb rehabilitation electromechanical devices that incorporate multimodal sensing and monitoring functions can significantly contribute to the recovery of limb motor functions. This report introduces a wearable mechanical pressure-electrophysiological monitoring bimodal sensing system specifically designed for human limb rehabilitation devices. By utilizing just four electrodes (SE, CE/DE, GND, REF), this system enables simultaneous and co-located measurement of surface electromyographic (sEMG), pressure, and mechanomyography (MMG) signals. These signals can be utilized to analyze muscle tension, stiffness, and tremor information. At last, this sensing system was used to assess muscle contraction force and localized muscle fatigue. The time and frequency domain characteristics of physiological signals during exercise were thoroughly investigated. The wearable mechanical pressure-electrophysiological bimodal sensing system can provide valuable data references for rehabilitation robots or human limb rehabilitation device, which is of great significance in the diagnosis of muscular diseases and rehabilitation treatment.

Communication between normal people and hearing-impaired people is usually realized with the sign language. However, the communication quality is limited by the familiarity of the people with sign language. Wearable sign language recognition systems, especially those based on flexible electronic glove, have attracted much attention due to their efficiency for sign language recognition. However, there are still two problems need to be solved in current research: 1) the size or weight of wearable device is usually inversely proportional to the fineness of the hand movement information captured and 2) both spatial features and temporal features are important for sign language recognition, and current algorithms usually do not achieve sufficient extraction. To address the above problems, a wearable electronic glove with new multilayer parallel LSTM-CNN (Para-LSTM-CNN) network is proposed. The designed wearable electronic glove is only 126 g in weight, and utilizes inertial measurement units (IMUs) and Flex sensors to capture hand motion signals. 41-dimensions of hand motion signals can be captured and uploaded to computer via Wi-Fi. A Para-LSTM-CNN algorithm based on parallel features extraction strategy is proposed. The temporal and spatial features of the sensor signals are extracted through long short-term memory (LSTM) and convolutional neural network (CNN), respectively. The Para-LSTM-CNN can effectively improve the accuracy of sign language recognition by fusing spatial and temporal features synchronously. The proposed system has a high sign-language-recognition accuracy of 97.01% for 26 gestures. Result shows that the proposed system promises to provide effective solutions for patients with hearing-impaired. The available code of Para-LSTM-CNN can be found athttps://github.com/1104162390-A/Para-LSTM-CNN/tree/main.

The significance of telerobotic trajectory segmentation has been demonstrated on a range of skill training and robotic automation tasks. However, the trajectories of telerobot are characterized by complexity and high dimensionality, making it difficult to segment them accurately. Existing methods are often plagued by feature inefficiency, and the clustering methods are also difficult to properly model the trajectories. In addition, over-segmentation also affects the accuracy of the clustering-based segmentation methods. To address above problems, this article presents a new unsupervised approach that automatically segments multimodal trajectory data of a telerobot and solves the problem of over-segmentation through a postpromoting procedure. First, we present an unsupervised visual feature-extraction network, that is, a dense connection spatial convolution network (DCSC) to generate more discriminative features for clustering. The dense convolution and spatial convolution facilitates information flow, enhances feature propagation, and avoids manual annotation. Next, we develop an unsupervised trajectory segmentation method that is called multimodality clustering with Chinese restaurant process (MC-CRP). The model utilizes data from different modalities and segments the trajectory through hierarchical clustering. MC-CRP obtains more accurate results in a short period of time. To further improve the precision of trajectory segmentation, we merge over-segments based on predefined similarity measurements. Extensive experiments on the publicly available data set JIGSAWS show that the presented approach achieves 70.1% silhouette coefficient, 25.1% normalized mutual information, and 71.4% segmentation accuracy. These metrics demonstrate that the presented segmentation approach provides deeper insight into the trajectory features and improve the accuracy of segmentation more efficiently than others.

Soft exosuits feature nonlinear, low stiffness, and hysteretic behavior, presenting different mechanical responses during loading and unloading of assistive force. Delivering the desired force to the wearer is a challenge in such a system. To address this issue, this article proposed a novel control strategy with adaptive stiffness and iterative position to optimize the assistive forces of the actuators in compliance with the humanexosuit interface stiffness and relative motion. Specifically, a stiffness model was proposed to describe the force loading and unloading behaviors, further an adaptive stiffness-based force controller was designed to improve the force tracking for whole profile by adaptively adjusting the stiffness parameters related to loading and unloading, and iteratively compensating the position error associated with the force amplitude in the model on a step-by-step basis. This control strategy was implemented on a soft exosuit for hip flexion and extension assistance, and its performance was evaluated through walking tests on six subjects. The results showed that after 4 or 5 iterations of optimization based on the initial parameter settings, the proposed controller could comply with the human-exosuit interface stiffness and achieved an improved force tracking, with a minimal root-mean-square error of 7.5 N in desired force profiles tracking, and a metabolic gain of 14.8%, demonstrating a promising potential of the proposed controller for improving the force tracking and the economy of human walking.Note to Practitioners—Soft exosuits have shown promising outcomes in augmenting human walking and reducing fatigue. However, delivering the desired assistive force accurately to the wearer remains challenging due to the nonlinear and variable stiffness nature of the human-exosuit interaction during walking. This article proposed a control strategy for improving the force tracking performance by optimizing the force-positional relationship of the actuators. Specifically, a stiffness model was established to define the relationship between the force and position of the actuators. This model acts as a virtual spring between the wearer and the exosuit, and stiffness of the spring was adjusted to adapt the wearer. On the Basis of this model, we developed a force tracking controller with adaptive stiffness. It can transmit accurately a desired force trajectory to the wearer by adaptively optimizing the stiffness parameters and iteratively compensating the position error in stiffness model. This controller was implemented on a soft exosuit for hip flexion and extension assistance. The results of treadmill and outdoor walking tests with six subjects confirm the significant effect on optimizing the force-positional relationship, improving the force tracking performance, as well as avoiding force hystereses or friction losses. This research addresses the challenge of force control in human-exosuit interaction with variable stiffness characteristics. The limitation of this work is that it only focuses on hip joint assistance. In the future, we hope to extend this control strategy to other joints to achieve the generality of the approach.

Pulse wave analysis, a non-invasive and cuff-less approach, holds promise for blood pressure (BP) measurement in precision medicine. In recent years, pulse wave learning for BP estimation has undergone extensive scrutiny. However, prevailing methods still encounter challenges in grasping comprehensive features from pulse waves and generalizing these insights for precise BP estimation. In this study, we propose a general pulse wave deep learning (PWDL) approach for BP estimation, introduc-ing the OVAR-BPnet model to powerfully capture intricate pulse wave features and showcasing its effectiveness on multiple types of pulse waves. The approach involves constructing population pulse waves and employing a model comprising an omni-scale convolution subnet, a Vision Transformer subnet, and a multilayer perceptron subnet. This design enables the learning of both single-period and multi-period waveform features from multiple subjects. Additionally, the approach employs a data augmentation strategy to enhance the morphological features of pulse waves and devise a label sequence regularization strategy to strengthen the intrinsic relationship of the subnets' output. Notably, this is the first study to validate the performance of the deep learning approach of BP estimation on three types of pulse waves: photoplethysmography, forehead imaging photoplethysmography, and radial artery pulse pressure waveform. Experiments show that the OVAR-BPnet model has achieved advanced levels in both evaluation indicators and international evaluation criteria, demonstrating its excellent competitiveness and generalizability. The PWDL approach has the potential for widespread application in convenient and continuous BP monitoring systems.

Wearable exoskeletons can help people with spinal cord injuries regain mobility. For the control of wearable exoskeletons, transient performance, and steady-state performance are the important indices to be considered. In this article, a model-free fractional-order prescribed performance sliding-mode control scheme is proposed for the joint angle tracking control. Time-delay control is developed based on the ultralocal model of the wearable exoskeleton to form the model-free feature. Afterward, a novel exponential-type barrier Lyapunov function is de ed to ensure the prescribed transient performance on the tracking errors. Then, a fractional-order sliding-mode surface is introduced, based on the exponential-type barrier Lyapunov function and fractional-order sliding-mode surface, a fixed-time sliding-mode controller is designed to further enhance the control performance. Benefiting from the abovementioned methods, the proposed controller is model-free and has fixed-time convergence with the prescribed performance. In addition, the adjustable range of the parameters is enlarged by the fractional-order sliding-mode surface. Stability is analyzed based on the Lyapunov theory. Finally, experiments are implemented in the wearable exoskeleton experimental platform, experiment results demonstrate the effectiveness of the proposed scheme.

Humanoid lower limbs with tactile cognition are crucial for future bipedal robots developing advanced bionic intelligence, such as owning autonomous reflexes and performing human-like actions. Most existing robotic lower limbs focus on providing physical support and mobility, with little work on more bionic DoFs or tactile sensing abilities that are more than significant for a fully humanoid system. This paper develops a robotic lower limb with whole-foot tactile sensing capability. An eight-DoF mechanism with humanoid joints is designed comprehensively, and a tactile sensor with electric double-layer capacitors principle wraps the special-shaped foot surface. An STM32-based circuit integrating perception and control with a real-time inverse kinematics algorithm is experimentally demonstrated. This work provides novel insight and methodology for humanoid robots and tactile-based bionic intelligence.

Reinforcement Learning (RL) with Skill Diversity (SD) is an appealing method of imitating expert trajectory. Numeral papers have presented methods of SD in several imitation learning scenarios. However, it is still a gap between SD and the practical implementation of robot trajectory imitation. The poor performance of matching algorithm often leads to imitation failures as well. To bridge the gaps and remedy the drawbacks, this paper proposed a robotic trajectory imitation method SD-PDMD with no-reward function reinforcement learning. A RL frame for robot trajectory imitation with SD is constructed to generate potential imitation trajectories, and then an optimal similarity algorithm based on predefined similarity measurements (PDMD) is proposed to match the expert trajectory with the most similar generated trajectory. Extensive experiments illustrate that the proposed SD-PDMD can effectively complete the robot trajectory imitation task, the performance of PDMD for similarity matching is also better than the traditional Euclidean Distance with an improvement of 5.7%-9.2%. The code and video can be obtained at: https://github.com/Nursing-Robot-Laboratory/SD-PDMD

Multi-size object detection is a technical difficulty which impeding the development of the intelligent nursing robot. To cope with the problem, this paper proposes a Pyramid Transformer model to detect the objects with different sizes in nursing scenario. Pyramid Transformer consists of three parts including Transformer Module, Pyramid Structure and Convolution Module. Transformer Module can improve the performance of large object detection with Multi-head Attention mechanism, and Pyramid Structure enables the model to make prediction with feature maps of different sizes which benefits the detection of small objects. Convolution Module is employed to reduce hardware requirements, and it makes Pyramid Transformer could run and implement on a single graphics card. The experiments show that the mean average precision reaches 72.7% which makes improvement over other models. This shows that the proposed Pyramid Transformer model is practical and effective for object detection of the nursing robot. The dataset can be got at https://github.com/NotFar1997/NSI-dataset.

To cope with the difficulty of training and eval-uation for nursing telerobot operator. This paper proposes a training-evaluation method for operator with unsupervised trajectory segmentation. To evaluate the dexterity and proce-dural knowledge of the operators objectively, we propose a new unsupervised model TSC-CRP that can automatically segment trajectory from nursing robotic training sessions. By comparing the segmented sub-trajectories and the standard sub-trajectory process, the method can provide objective evaluation and meaningful feedback without the intervention from experts. Experiments show that TSC-CRP has higher segmentation accuracy than other unsupervised methods, and it can identify the operators with different skill levels. In practical, the proposed training-evaluation system allows to provide an in-depth analysis of operator action to assess their skills precisely.

Regular maintenance of pipelines is an important task to ensure oil transportation and other operation (sewers, nature gas). Precise localization of pipeline damage can greatly improve the efficiency of maintenance work. Since the texture similarity and illumination change of pipe, traditional local descriptors for image matching like SIFT, SURF and ORB are easy to suffer from false correspondences. As to remove the false matches, the local neighborhood constraints (LNC) that contain spatial constructs around feature points are proposed. Good correspondences are essential for the high-accuracy localization and mapping solution given the limited textures and illumination in the pipes. The LNC method is also integrated into the state-of-the-art visual SLAM system. The proposed LNC image matching method and the SLAM system are evaluated on datasets gathered from the pipe environment. Compared with other state-of-the-art methods, our LNC image matching method achieves similar or better performance in precision, recall and runtime. The SLAM system provides state estimation and map reconstruction of the pipe in real-time, and the localization error is within 1%.

To cope with the requirements on efficiency and labour-saving of the out-pipe surface grinding tasks in the wild, several proposals are revealed and discussed. The one benefiting from the characteristics of planetary gear transmission and friction actuating mechanism are expatiated. To realize full coverage of out-pipe surface, the self-rotation and revolution motions of every polishing tool (cutter) are actuated by the same motor, and the friction force produced in grinding process acts as suitable tractive force for the forward travel of the grinding system. The friction statics analysis is established to illustrate the force transmission. Compression spring system are utilized to realize force equilibrium and support passive diameter adaptability. The proposed robotic grinding system is characterized by less actuator, online grinding capability and high working efficiency. It has clear advantages regarding manufacturing costs and control complexity. As the result of prototype experiments, performance of smooth grinding the out-pipe surface is confirmed.

Human touching the robot to convey intentions or emotions is an essential communication pathway during physical Human-Robot Interaction (pHRI). Therefore, advanced service robots require superior tactile intelligence to guarantee naturalness and safety when making physical contact with human subjects. Tactile intelligence is the capability to percept and recognize tactile information from touch behaviors, in which understanding the physical meaning of touching actions is crucial. For this purpose, this report introduces a recently collected and organized dataset "TacAct" that encloses real-time tactile information when human subjects touched the test device mimicking a robot forearm. The dataset contains 12 types of 24,000 touch actions from 50 subjects. The dataset details are described, the data are preliminarily analyzed, and the validity of the dataset is tested through a convolutional neural network LeNet-5 which classifying different types of touch actions. We believe that the TacAct dataset would be beneficial for the community to understand the touch intention under various circumstances and to develop learning-based intelligent algorithms for different applications.

This paper proposes a relatively simple function of assistive force for a belt-type hip assist suit developed by the authors' group. The function, which is inspired by the muscle force of the rectus femoris, contains only two parameters, the magnitude and a phase shift factor. Thus, it can reduce the amount of calculation in generating the desired assistive force during walking. Tests were performed on three healthy subjects to confirm its effect and to investigate its influence on the motions of hip, knee and ankle joints. It was demonstrated that the effect of the assist depended greatly on the phase shift factor, i.e., the location of the peak of the assistive force in a swing period. A large effect was observed when the peak of the assistive force came at mid-swing phase. The results of the tests showed that the proposed force function could help to increase walk ratio (the ratio of step length to the number of steps per minute) by an average value of 11.2% at a force magnitude of 35 N, which could produce an assistive torque of the same order as the magnitude of the muscle force of the rectus femoris around the hip joint.

In aging societies, there is a strong demand for robotics to tackle problems resulting from the aging population. Patient transfer, such as lifting and moving a bedridden patient from a bed to a wheelchair and back, is one of the most physically challenging tasks in nursing care. We have developed a prototype nursing-care assistant robot, RIBA, that can conduct patient transfer using human-type arms. The basic robot motion trajectories are created by interpolating several postures designated in advance. To accomplish more flexible and suitable motion, adjustment using sensor information is necessary, because the patient's posture and positions in contact with the robot may differ slightly in each trial. In this paper, we propose a motion adjustment method in patient lifting using tactile sensors mounted on the robot arms. The results of experiments using a lifesize dummy are also presented.

In aging societies, there is a strong demand for robotics to tackle problems resulting from the aging population. We have developed a prototype nursing-care assistant robot, RIBA, which was designed to come in direct contact with patients and conduct physically challenging tasks. RIBA interacts with its object, typically a human, through multiple and distributed contact regions on its arms and body. To obtain information on such whole-body contact, RIBA has tactile sensors on a wide area of its arms. The regions where hard contact with the manipulated person may occur have almost flat surfaces, leading to surface contact involving a finite area, in order to reduce contact pressure and not to cause the person's pain. When controlling the position and orientation of the person, the relative positions and orientations of the distributed contacting surfaces should be preserved as far as possible to maintain stable contact and not to graze the person's skin. Preserving the force and the pressure pattern of each contact region using tactile feedback is also important to provide stable and comfortable human-robot physical interaction. In this paper, we propose a whole-body contact manipulation method using tactile information to meet these requirements.

In aging societies, there is a strong demand for robotics to tackle problems caused by the aging population. Patient transfer, such as lifting and moving a bedridden patient from a bed to a wheelchair and back, is one of the most physically challenging tasks in nursing care, the burden of which should be reduced by the introduction of robot technologies. We have developed a new prototype robot named RIBA with human-type arms that is designed to perform heavy physical tasks requiring human contact, and we succeeded in transferring a human from a bed to a wheelchair and back. To use RIBA in changeable and realistic environments, cooperation between the caregiver and the robot is required. The caregiver takes responsibility for monitoring the environment and determining suitable actions, while the robot undertakes hard physical tasks. The instructions can be intuitively given by the caregiver to RIBA through tactile sensors using a newly proposed method named tactile guidance. In the present paper, we describe RIBA's design concept, its basic specifications, and the tactile guidance method. Experiments including the transfer of humans are also reported.