Search

Abstract The open‐circuit magnetic field of the V‐shaped interior PM (IPM) machine is predicted by the analytical method. The method is implemented following the basic principle of the subdomain method. The machine model is divided into five different regions, including rotor slot, magnetic barrier, air gap, stator slot, and stator slot opening. To perform the analytical prediction process in the polar coordinate system, these regions should be transformed into the appropriate subdomains. The last four regions can be transformed into the subdomains with the required shape based on the existing method. The rotor slot cannot be treated as a single subdomain, for the shape of which is irregular. Herein, the PM and vacuum region in the rotor slot is divided into several fan‐shaped subdomains. The division method is described clearly. Then, the boundary conditions alongside the interface between adjacent subdomains are derived based on function transformation. On this basis, the open‐circuit magnetic field of the IPM machine can be predicted. The magnetic field of the 6p/36 s V‐shaped IPM machine is obtained by the proposed method, finite element method simulation, and experiment, respectively. Additionally, the comparison results verify the accuracy and efficiency of the proposed method.

Accurate robot dynamics models are crucial for safe and stable control as well as for generalization to new conditions. Data-driven methods are increasingly used in robotics dynamics modeling for their superior approximation, with extrapolation performance being a critical efficacy indicator. While deep learning is widely used, it often overlooks essential physical principles, leading to weaker extrapolation capabilities. Recent innovations have introduced physics-inspired deep networks that integrate deep learning with physics, leading to improved extrapolation due to their informed structure, but potentially to underfitting in real-world scenarios due to the presence of unmodeled phenomena. This paper presents an experimental framework to assess the extrapolation capabilities of data-driven methods. Using this framework, physics-inspired deep networks are applied to learn the inverse dynamics models of a simulated robotic manipulator and two real physical systems. The results show that under ideal observation conditions physics-inspired models can learn the system’s underlying structure and demonstrate strong extrapolation capabilities, indicating a promising direction in robotics by offering more accurate and interpretable models. However, in real systems their extrapolation often falls short because the physical priors do not capture all dynamic phenomena, indicating room for improvement in practical applications.

Representing and rendering dynamic underwater scenes present significant challenges due to the medium’s inherent properties, which result in image blurring and information ambiguity. To overcome these challenges and accomplish real-time rendering of dynamic underwater environments while maintaining efficient training and storage, we propose Underwater Dynamic Scene Reconstruction Gaussian Splatting (UDR-GS), a method based on Gaussian Splatting. By leveraging prior information from a pre-trained depth estimation model and smoothness constraints between adjacent images, our approach uses the estimated depth as a geometric prior to aid in color-based optimization, significantly reducing artifacts and improving geometric accuracy. By integrating depth guidance into the Gaussian Splatting (GS) optimization process, we achieve more precise geometric estimations. To ensure higher stability, smoothness constraints are applied between adjacent images, maintaining consistent depth for neighboring 3D points in the absence of boundary conditions. The symmetry concept is inherently applied in our method by maintaining uniform depth and color information across multiple viewpoints, which enhances the reconstruction quality and visual coherence. Using 4D Gaussian Splatting (4DGS) as a baseline, our strategy demonstrates superior performance in both RGB novel view synthesis and 3D geometric reconstruction. On average, across multiple datasets, our method shows an improvement of approximately 1.41% in PSNR and a 0.75% increase in SSIM compared with the baseline 4DGS method, significantly enhancing the visual quality and geometric fidelity of dynamic underwater scenes.

Aiming at the problems of the poor robustness and universality of traditional contour matching algorithms in engineering applications, a method for improving the surface defect detection of industrial products based on contour matching algorithms is detailed in this paper. Based on the image pyramid optimization method, a three-level matching method is designed, which can quickly obtain the candidate pose of the target contour at the top of the image pyramid, combining the integral graph and the integration graph acceleration strategy based on weak classification. It can quickly obtain the rough positioning and rough angle of the target contour, which greatly improves the performance of the algorithm. In addition, to solve the problem that a large number of duplicate candidate points will be generated when the target candidate points are expanded, a method to obtain the optimal candidate points in the neighborhood of the target candidate points is designed, which can guarantee the matching accuracy and greatly reduce the calculation amount. In order to verify the effectiveness of the algorithm, functional test experiments were designed for template building function and contour matching function, including uniform illumination condition, nonlinear condition and contour matching detection under different conditions. The results show that: (1) Under uniform illumination conditions, the detection accuracy can be maintained at about 93%. (2) Under nonlinear illumination conditions, the detection accuracy can be maintained at about 91.84%. (3) When there is an external interference source, there will be a false detection or no detection, and the overall defect detection rate remains above 94%. It is verified that the proposed method can meet the application requirements of common defect detection, and has good robustness and meets the expected functional requirements of the algorithm, providing a strong technical guarantee and data support for the design of embedded image sensors in the later stage.

Artificial intelligence promotes the development of mineral/rock automatic recognition technology to reduce labor costs and reliance on personal experience. With instrument digitization, image recognition has played an increasingly important role. This review focuses on the image sensing system applied to mineral/rock identification and redivides this system into a sensing module, an imaging module, and an information processing module. The sensing and imaging mechanisms are summarized via technologies of sensing, imaging, and display. The information processing is classified into single‐modal or multimodal recognition. This review analyzes the limitations of single‐modal recognition and discusses the advantages and significance of multimodal recognition. The challenges of current mineral/rock identification are concluded, such as public database, multimodal learning, and portable sensing system. The interdisciplinary sensing technologies are discussed, and application feasibility is explored. This review aims to provide a reference for the research community in automatic mineral/rock identification and sensing system optimization.

Hitherto, automated grasping with robotic grippers requires adjusting the posture and force of the fingers based on the size, geometry, stiffness, and pose of the objects. To provide a simpler but efficient grasping methodology, a soft enveloping gripper is presented and investigated how its morphological adaptability improves the grasping ability by comparing its performance with fingered grippers. Results show that this enveloping gripper can omnidirectionally envelop objects via active–passive interaction, which allows the gripper to 100% grasp the object located at different positions within range and keep their orientations. However, the grasping success rate and orientation error of the fingered grippers highly depend on the relative position and angle of the objects to the grippers, as well as the number of fingers. The dynamic vibration and decay time of the enveloping gripper when grasping a 500 g weight are, both, approximately one‐sixth of those of the two‐fingered gripper when grasping a 12.37 g cube. This enveloping gripper can automatically grasp objects (including deformable ones) lying in different poses without posture estimation and force control with a simple vision‐based automatic grasping method. The enveloping grasping method may open an avenue for simple, low cost yet powerful automatic grasping applications.

In the quest for intelligent robots, it is essential to enable them to understand tasks beyond mere manipulation. Achieving this requires a robust parsing mode that can be used to understand human cognition and semantics. However, the existing methods for task and motion planning lack generalization and interpretability, while robotic knowledge bases primarily focus on static manipulation objects, neglecting the dynamic tasks and skills. To address these limitations, we present a knowledge-based framework for hierarchically understanding various factors and knowledge types in robotic manipulation. Using this framework as a foundation, we collect a knowledge graph dataset describing manipulation tasks from text datasets and an external knowledge base with the assistance of large language models and construct the knowledge base. The reasoning tasks of entity alignment and link prediction are accomplished using a graph embedding method. A robot in real-world environments can infer new task execution plans based on experience and knowledge, thereby achieving manipulation skill transfer.

Soft robots are challenging to model and program. Non-specialists face non-negligible obstacles when working with soft robots to perform tasks. Here, the authors propose a method to interactively teach soft robots complex motions through flexible touchless and tactile multimodal sensors.

Abstract With the advance of smart material science, robotics is evolving from rigid robots to soft robots. Compared to rigid robots, soft robots can safely interact with the environment, easily navigate in unstructured fields, and be minimized to operate in narrow spaces, owning to the new actuation and sensing technologies developed by the smart materials. In the review, different actuation and sensing technologies based on different smart materials are analyzed and summarized. According to the driving or feedback signals, actuators are categorized into electrically responsive actuators, thermally responsive actuators, magnetically responsive actuators, and photoresponsive actuators; sensors are categorized into resistive sensors, capacitive sensors, magnetic sensors, and optical waveguide sensors. After introducing the principle and several robotic prototypes of some typical materials in each category of the actuators and sensors. The advantages and disadvantages of the actuators and sensors are compared based on the categories, and their potential applications in robotics are also presented.

The natural interaction between the prosthetic hand and the upper limb amputation patient is important and directly affects the rehabilitation effect and operation ability. Most previous studies only focused on the interaction of gestures but ignored the force levels. This paper proposes a simultaneous recognition method of gestures and forces for interaction with a prosthetic hand. The multitask classification algorithm based on a convolutional neural network (CNN) is designed to improve recognition efficiency and ensure recognition accuracy. The offline experimental results show that the algorithm proposed in this study outperforms other methods in both training speed and accuracy. To prove the effectiveness of the proposed method, a myoelectric prosthetic hand integrated with tactile sensors is developed, and surface electromyography (sEMG) datasets of healthy persons and amputees are built. The online experimental results show that the amputee can control the prosthetic hand to continuously make gestures under different force levels, and the effect of hand coordination on the hand perception of amputees is explored. The results show that gesture classification operation tasks with different force levels based on sEMG signals can be accurately recognized and comfortably interact with prosthetic hands in real time. It improves the amputees’ operation ability and relieves their muscle fatigue.

Tactile object recognition is crucial for effective grasping and manipulation. Recently, it has started to attract increasing attention in robotic applications. While there are many works on tactile object recognition and they also achieved promising performances in some applications, most of them are usually limited to closed world scenarios, where the object instances to be recognition in deployment are known and the same as that of during training. Since robots usually operate in realistic open-set scenarios, they inevitably encounter unknown objects. If automation systems falsely recognize unknown objects as one of the known classes based on the pre-trained model, it can lead to potentially catastrophic consequences. It motivates us to break the closed world assumption and to study tactile object recognition in realistic open-set conditions. Although several open-set recognition methods have been proposed, they focused on visual tasks and may not be suitable for tactile recognition. It is mainly due to that these methods do not take into account the special characteristic of tactile data in their models. To this end, we develop a novel Gaussian Prototype Learning method for robust tactile object recognition. Particularly, the proposed method converts feature distributions to probabilistic representations, and exploit uncertainty for tactile recognition in open-set scenarios. Experiments on the two tactile recognition benchmarks demonstrate the effectiveness of the proposed method on open-set tasks.

Although intelligent technologies has facilitated the development of precise orthopaedic, simple internal fixation, ligament reconstruction or arthroplasty can only relieve pain of patients in short-term. To achieve the best recover of musculoskeletal injuries, three bottlenecks must be broken through, which includes scientific path planning, bioactive implants and personalized surgical channels building. As scientific surgical path can be planned and built by through AI technology, 4D printing technology can make more bioactive implants be manufactured, and variable structures can establish personalized channels precisely, it is possible to achieve satisfied and effective musculoskeletal injury recovery with the progress of multi-layer intelligent technologies (MLIT).

Myoelectric control for prosthetic hands is an important topic in the field of rehabilitation. Intuitive and intelligent myoelectric control can help amputees to regain upper limb function. However, current research efforts are primarily focused on developing rich myoelectric classifiers and biomimetic control methods, limiting prosthetic hand manipulation to simple grasping and releasing tasks, while rarely exploring complex daily tasks. In this article, we conduct a systematic review of recent achievements in two areas, namely, intention recognition research and control strategy research. Specifically, we focus on advanced methods for motion intention types, discrete motion classification, continuous motion estimation, unidirectional control, feedback control, and shared control. In addition, based on the above review, we analyze the challenges and opportunities for research directions of functionality-augmented prosthetic hands and user burden reduction, which can help overcome the limitations of current myoelectric control research and provide development prospects for future research.

Abstract Material recognition plays an important role in the interaction between robots and the external environment. For example, household service robots need to replace humans in the home environment to complete housework, so they need to interact with daily necessities and obtain their material performance. Images provide rich visual information about objects; however, it is often difficult to apply when objects are not visually distinct. In addition, tactile signals can be used to capture multiple characteristics of objects, such as texture, roughness, softness, and friction, which provides another crucial way for perception. How to effectively integrate multi‐modal information is an urgent problem to be addressed. Therefore, a multi‐modal material recognition framework CFBRL‐KCCA for target recognition tasks is proposed in the paper. The preliminary features of each model are extracted by cascading broad learning, which is combined with the kernel canonical correlation learning, considering the differences among different models of heterogeneous data. Finally, the open dataset of household objects is evaluated. The results demonstrate that the proposed fusion algorithm provides an effective strategy for material recognition.

Abstract Exoskeleton has been applied in the field of medical rehabilitation and assistance. However, there are still some problems in the interaction between human and exoskeleton, such as time delay, the existence of certain constraints on the human body, and the movement in time is hard to follow. A human motion pattern recognition model based on the long short‐term memory (LSTM) is proposed, which can recognise the state of the human body. Meanwhile, the orthogonalisation method is integrated to make personal‐specific disentangling, and it can effectively improve the generalisation ability of different groups of people, so as to improve the effective follower ability of the exoskeleton. Compared with some other traditional methods, this model has better performance and stronger generalisation ability, which has certain significance in the field of exoskeleton algorithm.

Autonomous indoor service robots are affected by multiple factors when they are directly involved in manipulation tasks in daily life, such as scenes, objects, and actions. It is of self-evident importance to properly parse these factors and interpret intentions according to human cognition and semantics. In this study, the design of a semantic representation framework based on a knowledge graph is presented, including (1) a multi-layer knowledge-representation model, (2) a multi-module knowledge-representation system, and (3) a method to extract manipulation knowledge from multiple sources of information. Moreover, with the aim of generating semantic representations of entities and relations in the knowledge base, a knowledge-graph-embedding method based on graph convolutional neural networks is proposed in order to provide high-precision predictions of factors in manipulation tasks. Through the prediction of action sequences via this embedding method, robots in real-world environments can be effectively guided by the knowledge framework to complete task planning and object-oriented transfer.

Sensory perception for dexterous robotic hands is an active research area and recent progress in robotics. Effective dexterous manipulation requires robotic hands to accurately feedback their state or perceive the surrounding environment. This article reviews the state-of-the-art of sensory perception for dexterous robotic manipulation. Two types of sensors, such as intrinsic and extrinsic sensors, are introduced according to their function and layout in robotic hands. These sensors provide rich information to a robotic hand, which contains the posture, the contact information of objects, and the physical information of the environment. Then, a comprehensive analysis of perception methods including planning-level, control-level, and learning-level perceptions is presented. The information obtained from sensory perception can help robotic hands to make decisions effectively. Previously issued reviews mainly focus on the design of tactile senor, while we analyze and discuss the relationship among sensing, perception, and dexterous manipulation. Some potential research topics on sensory perception are also summarized and discussed.

For higher effective volume ratio of roots pump， from high output production and low inner leakage， the shaped-profile configuration of a ellipse basic rotor and its raised rotor of is constructed. Firstly， from ellipse coordinate equation of rotor shaped-profile and conjugate principle of rotor pairs and critical conjugate conditions， the optimum axial length ratio of ellipse arc and the maximum shape coefficient of basic rotor are calculated. Secondly， based on the geometric relationship when the endpoint of valley working shaped-profile on driving rotor is just perfect to avoid the peak corner point on driven rotor， the maximum shape factor of raised rotor under a given addendum angle is determined. Finally， all parameter coordinate equations of raised shaped-profiles are provided， and the comparative advantage results of raised rotor to common circular arc rotor is focused on. All results show that the elliptical arc foundation rotor and high-profile rotor profile construction method have significant improvement in both high yield and leakage prevention effects， and the fundamental problem of the optimal axial length ratio and maximum shape factor of the arcs arc of the Roots pump is solved.

Abstract This paper investigates the new stability criteria for the asymptotic stability of time-delay systems via integral inequalities and Jensen inequalities. Firstly, not only the known constant time delay, but also the unknown time-varying delay is considered for the linear system. Secondly, the new delay-dependent Lyapunov–Krasovskii functional based on the double integral inequalities and Jensen inequalities is introduced, such that the linear system with time-delay is asymptotically stable. Thirdly, two classes of delay-dependent stability conditions in terms of linear matrix inequalities (LMIs) are derived, such that the control design conditions are relaxed and computation complexity is reduced. Compared with previous works, the larger feasible solution region and less conservative results are obtained. Finally, some numerical examples are performed to show the effectiveness and advantage of the proposed method.

In daily life, there are a variety of complex sound sources. It is important to effectively detect certain sounds in some situations. With the outbreak of COVID-19, it is necessary to distinguish the sound of coughing, to estimate suspected patients in the population. In this paper, we propose a method for cough recognition based on a Mel-spectrogram and a Convolutional Neural Network called the Cough Recognition Network (CRN), which can effectively distinguish cough sounds.

Robotic exoskeletons are developed with the aim of enhancing convenience and physical possibilities in daily life. However, at present, these devices lack sufficient synchronization with human movements. To optimize human-exoskeleton interaction, this article proposes a gait recognition and prediction model, called the gait neural network (GNN), which is based on the temporal convolutional network. It consists of an intermediate network, a target network, and a recognition and prediction model. The novel structure of the algorithm can make full use of the historical information from sensors. The performance of the GNN is evaluated based on the publicly available HuGaDB dataset, as well as on data collected by an inertial-based wearable motion capture device. The results show that the proposed approach is highly effective and achieves superior performance compared with existing methods.

A novel variable stiffness soft robotic hand (SRH) consists of three pieces of layer jamming structure (LJS) is proposed. The mechanism is driven by the motor-based tendon along the surface of the pieces that connect to individual gas channel. Each LJS is optimised by adhering a thin layer of hot melt adhesive and overlapping the spring steel sheet as inner layer material. It can be switched between rigid and compliant independently. The structures of variable stiffness and tendon-driven lead to various deformation poses. Then the control system of SRH and the performance analysis of the LJS are introduced. Finally, the experiments are implemented to prove the superiority of the proposed LJS and the demonstrations show that the designed robotic hand has multiple configurations to successfully grasp various objects.

Material recognition is an essential problem in the industrial automation community. In this study, we develop a machine learning method to identify the wood material. We extract various feature descriptors for the sound signal and perform a comprehensive comparison with the common classifier. The experimental validations achieve the best feature combination. Besides, we developed a practical stylus to collect sound signal from the wooden samples and present promising preliminary results.

Object recognition in containers is extremely difficult for robots. Dynamic audio signals are more responsive to an object's internal property. Therefore, we adopt the dynamic contact method to collect acoustic signals in the container and recognize objects in containers. Traditional machine learning is to recognize objects in a closed environment, which is not in line with practical applications. In real life, exploring objects is dynamically changing, so it is necessary to develop methods that can recognize all classes of objects in an open environment. A framework for recognizing objects in containers using acoustic signals in an open environment is proposed, and then the kernel k nearest neighbor algorithm in an open environment (OSKKNN) is set. An acoustic dataset is collected, and the feasibility of the method is verified on the dataset, which greatly promotes the recognition of objects in an open environment. And it also proves that the use of acoustic to recognize objects in containers has good value.

We present an attention based visual analysis framework to compute grasp-relevant information which helps to guide grasp planning using a multi-fingered robotic hand. Our approach uses a computational visual attention model to locate regions of interest in a scene and employ a deep convolutional neural network to detect grasp type and grasp attention point for a sub-region of the object in a region of interest. We demonstrate the proposed framework with object grasping tasks, in which the information generated from the proposed framework is used as prior information to guide grasp planning. The effectiveness of the proposed approach is evaluated in both simulation experiments and real-world experiments. Experimental results show that the proposed framework can not only speed up grasp planning with more stable configurations, but also handle unknown objects. Furthermore, our framework can handle cluttered scenarios. A new Grasp Type Dataset (GTD) which includes six commonly used grasp types and covers 12 household objects is also presented.

The brain–computer interface-based rehabilitation robot has quickly become a very important research area due to its natural interaction. One of the most important problems in brain–computer interface is that large-scale annotated electroencephalography data sets required by advanced classifiers are almost impossible to acquire because biological data acquisition is challenging and quality annotation is costly. Transfer learning relaxes the hypothesis that the training data must be independent and identically distributed with the test data. It can be considered a powerful tool for solving the problem of insufficient training data. There are two basic issues with transfer learning, under transfer and negative transfer. We proposed a novel brain–computer interface framework by using autoencoder-based transfer learning, which includes three main components: an autoencoder framework, a joint adversarial network, and a regularized manifold constraint. The autoencoder framework automatically encodes and reconstructs data from source and target domains and forces the neural network to learn to represent these domains reliably. The joint adversarial network aims to force the network to learn to encode more appropriately for the source domain and target domain simultaneously, thereby overcoming the problem of under transfer. The regularized manifold constraint aims to avoid the problem of negative transfer by avoiding geometric manifold structure in the target domain being destroyed by the source domain. Experiments show that the brain–computer interface framework proposed by us can achieve better results than state-of-the-art approaches in electroencephalography signal classification tasks. This is helpful in aiding our rehabilitation robot to understand the intention of patients and can help patients to carry out rehabilitation exercises effectively.

In this study, the authors study a deep learning model that can convert vision into tactile information, so that different texture images can be fed back to the tactile signal close to the real tactile sensation after training and learning. This study focuses on the classification of different image visual information and its corresponding tactile feedback output mode. A training model of ensembled generative adversarial networks is proposed, which has the characteristics of simple training and stable efficiency of the result. At the same time, compared with the previous methods of judging the tactile output, in addition to subjective human perception, this study also provides an objective and quantitative evaluation system to verify the performance of the model. The experimental results show that the learning model can transform the visual information of the image into the tactile information, which is close to the real tactile sensation, and also verify the scientificity of the tactile evaluation method.

This article introduces a visual–tactile multimodal grasp data set, aiming to further the research on robotic manipulation. The data set was built by the novel designed dexterous robot hand, the Intel’s Eagle Shoal robot hand (Intel Labs China, Beijing, China). The data set contains 2550 sets data, including tactile, joint, time label, image, and RGB and depth video. With the integration of visual and tactile data, researchers could be able to better understand the grasping process and analyze the deeper grasping issues. In this article, the building process of the data set was introduced, as well as the data set composition. In order to evaluate the quality of data set, the tactile data were analyzed by short-time Fourier transform. The tactile data–based slip detection was realized by long short-term memory and contrasted with visual data. The experiments compared the long short-term memory with the traditional classifiers, and generalization ability on different grasp directions and different objects is implemented. The results have proved that the data set’s value in promoting research on robotic manipulation area showed the effective slip detection and generalization ability of long short-term memory. Further work on visual and tactile data will be devoted to in the future.

Compared with traditional rigid gripper with joint-linkage structure, novel soft robotic gripper gives rise to continuous concern for the advantages of no-damage grasping, convenient manufacture, easy control, and low cost. In this study, we design and built two kinds of soft robotic grippers with four fiber-reinforced soft actuators which are distributed in circular and rectangle shapes for single and twin contacts grasping. A novel hybrid valve pneumatic control scheme combining proportional and solenoid valves is proposed. Also, a mode controllable hybrid valve pressure control method is proposed to adjust internal pressure of soft robotic grippers to adapt to different grasping tasks. The experiment results verify that the performances of hybrid valve outperform those of individual proportional valve or solenoid valve in the aspects of response time and steady-state accuracy. The hybrid valve has wide range of pressure regulation, result in that the soft robotic grippers are qualified to grasp various objects with different shapes, sizes, and weights.

Conventional visual perception technology is subject to many restrictions, such as illumination, background clutter, and occlusion. Many intrinsic properties of objects, like stiffness, hardness, and internal state, cannot be effectively perceived by visual sensors. For robots, tactile perception is a key approach to obtain environmental and object information. Different from vision sensors, tactile sensors can directly measure some physical properties of objects and environment. At the same time, humans also utilize touch sensory receptors as an important means to perceive and interact with the environment. In this article, we present a detailed discussion on tactile object recognition problem. We divide the current studies on the tactile object recognition into three subcategories and detailed analysis has been put forward on them. In addition, we also discuss some advanced topics such as visual–tactile fusion, exploratory procedure, and data sets.

Teleoperation is of great importance in the area of robotics especially when people’s presence at the robot working space is unavailable. It provides an alternative to employ human intelligence in the control of the robot remotely. We establish robotic teleoperation systems with a wearable multimodal fusion device. The device is integrated with 18 low-cost inertial and magnetic measurement units, which cover all segments of the arm and hand. The multimodal fusion algorithm based on extended Kalman filter is deduced to determine the orientations and positions of each segment. Then, the robotic teleoperation systems using the proposed device are designed. The novel teleoperation schemes can be applied for 11DOF robotic arm–hand system and 10DOF robotic arm–hand system, in which the operator’s fingers are used for robotic hand teleoperation, and the arms with palm are used for robotic arm teleoperation. Meanwhile, the proposed robotic teleoperation systems are fully realized with a user-friendly human–machine interaction interface. Finally, a series of experiments are conducted with our robotic teleoperation system successfully.

Grasping has always been a great challenge for robots due to its lack of the ability to well understand the perceived sensing data. In this work, we propose an end-to-end deep vision network model to predict possible good grasps from real-world images in real time. In order to accelerate the speed of the grasp detection, reference rectangles are designed to suggest potential grasp locations and then refined to indicate robotic grasps in the image. With the proposed model, the graspable scores for each location in the image and the corresponding predicted grasp rectangles can be obtained in real time at a rate of 80 frames per second on a graphic processing unit. The model is evaluated on a real robot-collected data set and different reference rectangle settings are compared to yield the best detection performance. The experimental results demonstrate that the proposed approach can assist the robot to learn the graspable part of the object from the image in a fast manner.