Rong Xiong

Balance motion generation for a humanoid robot playing table tennis

Rapid advancement in the field of humanoid robot has been on-going in the last ten years, however, the common goal that robot can interact with human being to achieve a complicate task remains a difficult problem. We take table tennis playing as a demonstration to study how a humanoid robot shall interact with the environment and human beings and developed a humanoid robot ‘Kong’ that can play table tennis and rally with a human player continuously. This paper focuses on robots stability keeping during table tennis playing. This problem becomes crucial due to the recoil force resulted from large arm acceleration. We introduce an optimal momentum compensation method and a position-based impedance control scheme to generate body motion so that the robot absorb the recoil force and ensure successful hit to the ball. The experimental results demonstrate that the robot is capable to rally with a human player and can act against the recoil force to maintain stability.

Stereo Visual-Inertial Odometry With Multiple Kalman Filters Ensemble

In this paper, we present a stereo visual-inertial odometry algorithm assembled with three separated Kalman filters, i.e., attitude filter, orientation filter, and position filter. Our algorithm carries out the orientation and position estimation with three filters working on different fusion intervals, which can provide more robustness even when the visual odometry estimation fails. In our orientation estimation, we propose an improved indirect Kalman filter, which uses the orientation error space represented by unit quaternion as the state of the filter. The performance of the algorithm is demonstrated through extensive experimental results, including the benchmark KITTI datasets and some challenging datasets captured in a rough terrain campus.

Robust and Accurate Multiple-camera Pose Estimation Toward Robotic Applications

Pose estimation methods in robotics applications frequently suffer from inaccuracy due to a lack of correspondence and real-time constraints, and instability from a wide range of viewpoints, etc. In this paper, we present a novel approach for estimating the poses of all the cameras in a multi-camera system in which each camera is placed rigidly using only a few coplanar points simultaneously. Instead of solving the orientation and translation for the multi-camera system from the overlapping point correspondences among all the cameras directly, we employ homography, which can map image points with 3D coplanar-referenced points. In our method, we first establish the corresponding relations between each camera by their Euclidean geometries and optimize the homographies of the cameras; then, we solve the orientation and translation for the optimal homographies. The results from simulations and real case experiments show that our approach is accurate and robust for implementation in robotics applications. Finally, a practical implementation in a ping-pong robot is described in order to confirm the validity of our approach.

Fast segmentation and identification in vision system for soccer robots

Real-time segmentation is important in the color-vision system on the robot system such as soccer robot. This paper introduced a color image segmentation method using improved seed-fill algorithm in YUV color space. The new method dramatically reduces the work of calculation, and speed up the image processing. The result of comparing it with the old method based on RGB color space was showed in the paper. The implementation on MiroSot Soccer Robot System shows that the new method is fast and accurate.

Impedance Control and its Effects on a Humanoid Robot Playing Table Tennis

This paper proposes an impedance control scheme used on humanoid robots for stability maintenance when the robot is expected to carry out fast manipulatory tasks. We take table tennis playing as an example to study this issue. The fast acceleration required by table tennis rallying will result in an unknown large reaction force on the robot, causing the body to swing back and forth in an oscillating motion and the foot to lose complete contact with the ground. To improve the stability during fast manipulation and in order to resist disturbances due to the reaction force, we introduce impedance control to absorb the impact and decrease the amplitude of body swinging. The systems adjusting time is also reduced and the oscillations are eliminated according to the experimental results, which show the effectiveness of our scheme.

Facial feature points detecting based on Gaussian Mixture Models

We use Gaussian Mixture Models to approximate the confidence maps analytically.We also use Gaussian Mixture Models to model the distribution of face shape.The posterior is maximized by iteratively maximizing a lower bound of it.High-order geometry constraint within facial features are considered.Model based Hough Voting scheme is proposed to estimate an initial face shape. Detecting predefined facial feature points in a human face image is a well studied problem. Despite the impressive achievements that have been made, it is still open under unconstrained environments, with variations of illumination, expression, head pose, as well as partial occlusions. This paper proposed a novel method to locate facial feature points under the variations mentioned above. Support vector machines with probability outputs are trained to provide the observation probability of each facial feature point. The observation probabilities, as well as the distribution of face shape which serves as the prior to constrain the relative position of the facial feature points, are both approximated with Gaussian Mixture Models. The problem is solved by maximizing the posterior which combines the prior and observation probability within the framework of Bayesian Inference. An optimization algorithm is developed to maximize the posterior by iteratively maximizing the lower bound of it. The proposed method preserves the high-order geometric constraint within facial feature points. With a simple initialization method of Model based Hough Voting, the method shows competitive detecting rate and locating accuracy on the LFPW and LFW datasets, compared to the methods of state-of-the-art.

A humanoid robot for table tennis playing

Humanoid robot has been one of the most active research topics in the field of robotics. Their human-like form and configuration gives it advantages in working in human-interactive environment. The bipedal walking capability makes them possible to step over and onto obstacles, providing accessibility and mobility in cluttered space. The multi-DOF design of arms and legs enables them assist or replace humans in their normal tasks, making human life easier and safer. Humanoid robots, with their human-like outlook, also bring better interactive experience and are expect to play a part in peoples daily life and help the elderly and the children.

Adaptive Torque and Position Control for a Legged Robot Based on a Series Elastic Actuator

Inspired by biological systems, we seek to achieve natural dynamics and versatile locomotion for hopping or running robots by installing a series elastic actuator (SEA) in the joints due to its compliant property, passive adaptability and energy storage. However, robots equipped with these actuators have drawbacks in terms of substantial delay and limited bandwidth in their position control, especially when a robot has to choose its foothold while it is running at a demanding speed. To solve these problems, compliance control and adaptive position/torque control are introduced to a hopping- legged robot in this paper. The compliant performance of the robot can be improved through the intrinsic property of an SEA with a torque control algorithm. Combining the kinetics model and stochastic model of a 2-DOF robot, an adaptive position control with Kalman Filtering (KF) is developed to provide rapid convergent state estimation of the load on the robotic end-effector by solving the inverse dynamics. Validating the robustness and effectiveness of the proposed algorithm on our hopping-legged robot Tigger, the experimental results show very good position-tracking and disturbance-rejection, as well as flexible interactions while operating in a complex environment.

Towards learning from demonstration system for parts assembly: A graph based representation for knowledge

The industrial robot plays an increasingly important role in manufacturing industry, but it is limited by its high prerequisite of programming on users, which is time consuming and challenging. In this context, a mechanism enabling autoprogram of robot is desired. Learning from demonstration system (LDS) is one of systems that aim on this goal. In this paper, we present a graph based representation of knowledge called assembly graph (AG) to describe the knowledge on parts assembly, which is independent on specific workspace or robot. Based on this graph representation, a LDS framework is then proposed for parts assembly to enable knowledge transform and knowledge transfer. The former means that the knowledge on part assembly can be transformed from the human to the robot and the latter means that the knowledge can be transferred from the leaned computer to any other robots with other workspace. Besides, we present an implementation of this conceptual framework consisting of two stages including demonstration and execution. In demonstration stage, a human teacher shows the process of parts assembly, which is recorded by a camera system. The assembly relations are detected from images and represented by AG. It is solved by taking property of parts as well as the robot into consideration to obtain the precise pose of each part. By using the solution, the robot can assemble the parts as shown in demonstration in the execution stage. In the experiment, the building blocks are used for assembly. Our robot succeeds in assembling some building blocks together which is initially demonstrated by a human teacher.

Vibration suppression based on input shaping for biped walking

The non-rigid behavior of a humanoid robot system will cause vibrations during walking which can affect the walking stability. In this paper, a feed forward controller based on the input shaping method is proposed to reduce such vibrations, which are modeled with second-order processes. Then the corresponding input shapers are designed and applied to the reference ZMP trajectory. The motions generated from the modified referenced ZMP trajectory will yield significantly less vibrations. The effectiveness of our method was validated through experiments on a full-size humanoid robot.