Search results

In this work, the authors aim to provide a set of convenient methods for generating training data, and then develop a deep learning method based on point clouds to estimate the pose of target for robot grasping.

This work presents a deep learning method PointSimGrasp on point clouds for robot grasping. In PointSimGrasp, a point cloud emulator is introduced to generate training data and a pose estimation algorithm, which, based on deep learning, is designed. After trained with the emulation data set, the pose estimation algorithm could estimate the pose of target.

In experiment part, an experimental platform is built, which contains a six-axis industrial robot, a binocular structured-light sensor and a base platform with adjustable inclination. A data set that contains three subsets is set up on the experimental platform. After trained with the emulation data set, the PointSimGrasp is tested on the experimental data set, and an average translation error of about 2–3 mm and an average rotation error of about 2–5 degrees are obtained.

The contributions are as follows: first, a deep learning method on point clouds is proposed to estimate 6D pose of target; second, a convenient training method for pose estimation algorithm is presented and a point cloud emulator is introduced to generate training data; finally, an experimental platform is built, and the PointSimGrasp is tested on the platform.

Humanoids have become the center of attraction for many researchers dealing with robotics investigations by their ability to replace human efforts in critical interventions. As a result, navigation and path planning has emerged as one of the most promising area of research for humanoid models. In this paper, a fuzzy logic controller hybridized with genetic algorithm (GA) has been proposed for path planning of a humanoid robot to avoid obstacles present in a cluttered environment and reach the target location successfully. The paper aims to discuss these issues.

Here, sensor outputs for nearest obstacle distances and bearing angle of the humanoid are first fed as inputs to the fuzzy logic controller, and first turning angle (TA) is obtained as an intermediate output. In the second step, the first TA derived from the fuzzy logic controller is again supplied to the GA controller along with other inputs and second TA is obtained as the final output. The developed hybrid controller has been tested in a V-REP simulation platform, and the simulation results are verified in an experimental setup.

By implementation of the proposed hybrid controller, the humanoid has reached its defined target position successfully by avoiding the obstacles present in the arena both in simulation and experimental platforms. The results obtained from simulation and experimental platforms are compared in terms of path length and time taken with each other, and close agreements have been observed with minimal percentage of errors.

Humanoids are considered more efficient than their wheeled robotic forms by their ability to mimic human behavior. The current research deals with the development of a novel hybrid controller considering fuzzy logic and GA for navigational analysis of a humanoid robot. The developed control scheme has been tested in both simulation and real-time environments and proper agreements have been found between the results obtained from them. The proposed approach can also be applied to other humanoid forms and the technique can serve as a pioneer art in humanoid navigation.

This paper aims to understand user information behaviours when they perform multilingual information retrieval. It also offers reference for the development of multilingual information retrieval systems and relevant service platforms.

The authors designed an experiment on multilingual information retrieval with WorldWideScience, utilized Camtasia studio7 (a screen capturing and recording tool) to record overall operational processes of subjects and collected participants’ thought processes with think-aloud protocols. Meanwhile, a questionnaire survey and interviews were used to examine the subjects’ background information, their feelings for the experiment and their ideas about the experimental platform, respectively. Thirty-two valid data points were obtained by 41 subjects.

The users preferred their own language for retrieval. Most users from social science chose general search or advanced search freely according to the tasks. The majority of the participants selected key words directly from the tasks as search terms. Doctoral candidates were more likely to construct a search query with logic symbols. Translation tools were utilized for assisting retrieval and solving doubts of translation. When facing obstacles, users stayed on the original web page to explore continually, followed by back to homepage.

This paper provides a study of user behaviour through investigating how users behave on the whole process of retrieving multilingual information. The findings offer advice for optimizing the function of multilingual information retrieval systems and service platforms.

Deep learning (DL) technology is used to design a voice evaluation system to understand the impact of learning aids on DL and mobile platforms on students’ learning behavior.

DL technology is used to design a speech evaluation system.

The experimental results show that the speech evaluation system designed has a high accuracy rate, the highest agreement rate with manual evaluation of pronunciation is 89.5%, and the correct speech recognition rate is 96.64%. The designed voice evaluation system and the manual voice rating system have a maximum error rate of 2%. The experimental results suggest that it is necessary to further optimize the learning aids for mobile platform. The learning aids of the mobile platform need to be further optimized to promote the improvement of student learning efficiency.

The results show that the speech evaluation system designed has good practical application value, and it provides a certain reference value for the future study of learning tools on DL.

The purpose of this paper is to investigate the usability (effectiveness, efficiency and user satisfaction) of e‐feedback interfaces. The experiment compares a traditional visual approach with a multimodal approach in order to determine the impact of multimodal metaphors upon the user's understanding, reasoning and engagement with the e‐feedback.

The empirical investigation involved visual (text with graphical illustrations) and multimodal (audio‐visual with expressive avatars and recorded speech) experimental e‐feedback platforms. Both experimental platforms provided the same e‐feedback but used different interaction metaphors to convey the information. The evaluation approach measured effectiveness, efficiency and user satisfaction.

The results showed that the multimodal approach increased usability in terms of effectiveness, efficiency and engagement of users with the e‐feedback. There is a very clear prima facie case that combining different communication metaphors to convey information involved in the e‐feedback simultaneously does not increase the information overload on users. This however was observed to be the case when the visual channel was used.

This paper introduces a unique approach that uses specific combinations of multimodal metaphors to communicate information about e‐feedback simultaneously. This approach increased the usability of e‐feedback and user's engagement in interfaces for e‐learning applications.

The purpose of this study is to examine the process underlying how consumers shift roles from users to suppliers of goods or services in the collaborative economy (CE). It examines quantatively the impact of a series of explanatory variables underlying that switchover process.

This study identifies and tests the key factors that motivate the user-provider transition by introducing the spillover effect from the proenvironmental literature into collaborative practices and using four experimental designs. Considering behavioral characteristics, context, intrinsic variables and socialization, this study provides an in-depth understanding of the process of transition from user to supplier in the CE.

The results suggest the interactive nature of the spillover as peer influence boosts changes in individual motivations, preferences and behaviors. Furthermore, promoting solidarity between members of the CE platform facilitates the transition of participants from users to providers. In addition, the users’ perception of socialization, satisfaction and sense of indebtedness may also play a significant role in the transition.

The study highlights the process underlying the switchover from user to provider at the prosumer level. More specifically, this study identifies key variables influencing the intention to switchover in the CE by drawing on the spillover effect from pro-environmental behavior and considering the spillover as an interactive process.

Managers who wish to develop collaborative systems must attract a critical mass of providers to ensure the viability of their systems. Instead of recruiting new providers, managers may convert existing users into providers. This study identifies the key variables to modulate to this end.

The findings offer important managerial implications and shed new light on the CE literature.

Many metal workpieces have the characteristics of less texture, symmetry and reflectivity, which presents a challenge to existing pose estimation methods. The purpose of this paper is to propose a pose estimation method for grasping metal workpieces by industrial robots.

Dual-hypothesis robust point matching registration network (RPM-Net) is proposed to estimate pose from point cloud. The proposed method uses the Point Cloud Library (PCL) to segment workpiece point cloud from scenes and a trained-well robust point matching registration network to estimate pose through dual-hypothesis point cloud registration.

In the experiment section, an experimental platform is built, which contains a six-axis industrial robot, a binocular structured-light sensor. A data set that contains three subsets is set up on the experimental platform. After training with the emulation data set, the dual-hypothesis RPM-Net is tested on the experimental data set, and the success rates of the three real data sets are 94.0%, 92.0% and 96.0%, respectively.

The contributions are as follows: first, dual-hypothesis RPM-Net is proposed which can realize the pose estimation of discrete and less-textured metal workpieces from point cloud, and second, a method of making training data sets is proposed using only CAD models with the visualization algorithm of the PCL.

This paper aims to present a new method to analyze the robot’s obstacle negotiation based on the terramechanics, where the terrain physical parameters, the sinkage and the slippage of the robot are taken into account, to enhance the robot’s trafficability.

In this paper, terramechanics is used in motion planning for all-terrain obstacle negotiation. First, wheel/track-terrain interaction models are established and used to analyze traction performances in different locomotion modes of the reconfigurable robot. Next, several key steps of obstacle-climbing are reanalyzed and the sinkage, the slippage and the drawbar pull are obtained by the models in these steps. In addition, an obstacle negotiation analysis method on loose soil is proposed. Finally, experiments in different locomotion modes are conducted and the results demonstrate that the model is more suitable for practical applications than the center of gravity (CoG) kinematic model.

Using the traction performance experimental platform, the relationships between the drawbar pull and the slippage in different locomotion modes are obtained, and then the traction performances are obtained. The experimental results show that the relationships obtained by the models are in good agreement with the measured. The obstacle-climbing experiments are carried out to confirm the availability of the method, and the experimental results demonstrate that the model is more suitable for practical applications than the CoG kinematic model.

Comparing with the results without considering Terramechanics, obstacle-negotiation analysis based on the proposed track-terrain interaction model considering Terramechanics is much more accurate than without considering Terramechanics.

This paper aims to study a disturbance rejection controller to improve the anti-interference capability and the position tracking performance of airborne radar stabilized platform that ensures the stability and clarity of synthetic aperture radar imaging.

This study proposes a disturbance rejection control scheme for an airborne radar stabilized platform based on the active disturbance rejection control (ADRC) inverse estimation algorithm. Exploiting the extended state observer (ESO) characteristic, an inversely ESO is developed to inverse estimate the unmodeled state and extended state of the platform system known as total disturbances, which greatly improves the estimation performance of the disturbance. Then, based on the inverse ESO result, feedback the difference between the output of the tracking differentiator and the inverse ESO result to the nonlinear state error feedback controller (NLSEF) to eliminate the effects of total disturbance and ensure the stability of the airborne radar stabilized platform.

Simulation experiments are adopted to compare the performance of the ADRC inverse estimation algorithm with that of the proportional integral derivative controller which is one of the mostly applied control schemes in platform systems. In addition, classical ADRC is compared as well. The results have shown that the ADRC inverse estimation algorithm has a better disturbance rejection performance when disturbance acts in airborne radar stabilized platform, especially disturbed by continuous airflow under some harsh air conditions.

The originality of this paper is exploiting the ESO characteristic to develop an inverse ESO, which greatly improves the estimation performance of the disturbance. And the ADRC inverse estimation algorithm is applied to ameliorate the anti-interference ability of the airborne radar stabilization platform, especially the ability to suppress continuous interference under complex air conditions.

This paper aims to propose an odor sensing system based on wide spectrum for e-nose, based on comprehensive analysis on the merits and drawbacks of current e-nose.

The wide spectral light is used as the sensing medium in the e-nose system based on continuous wide spectrum (CWS) odor sensing, and the sensing response of each sensing element is the change of light intensity distribution.

Experimental results not only verify the feasibility and effectiveness of the proposed system but also show the effectiveness of least square support vector machine (LSSVM) in eliminating system errors.

Theoretical model of the system was constructed, and experimental tests were carried out by using NO₂ and SO₂. System errors in the test data were eliminated using the LSSVM, and the preprocessed data were classified by euclidean distance to centroids (EDC), k-nearest neighbor (KNN), support vector machine (SVM), LSSVM, respectively.

The system not only has the advantages of current e-nose but also realizes expansion of sensing array by means of light source and the spectrometer with their wide spectrum, high resolution characteristics which improve the detection accuracy and realize real-time detection.