Spin Detection in Robotic Table Tennis*

  title={Spin Detection in Robotic Table Tennis*},
  author={Jonas Tebbe and Lukas Klamt and Andreas Zell},
  journal={2020 IEEE International Conference on Robotics and Automation (ICRA)},
In table tennis, the rotation (spin) of the ball plays a crucial role. A table tennis match will feature a variety of strokes. Each generates different amounts and types of spin. To develop a robot that can compete with a human player, the robot needs to detect spin, so it can plan an appropriate return stroke. In this paper we compare three methods to estimate spin. The first two approaches use a high-speed camera that captures the ball in flight at a frame rate of 380 Hz. This camera allows… 
Research on real-time evaluation algorithm of human movement in tennis training robot
Another approach to perform programmed ID of tennis swings with ball impact error technology relies on movement slope vector stream and polynomial relapse, and RBF classifiers can recognize beforehand undetectable bogus swings.
Optimal Stroke Learning with Policy Gradient Approach for Robotic Table Tennis
A realistic simulation environment where several models are built for the ball's dynamics and the robot’s kinematics is proposed, and an efficient retraining method is developed that significantly outperforms the existing RL methods in simulation.
Sample-efficient Reinforcement Learning in Robotic Table Tennis
This paper presents a sample-efficient RL algorithm applied to the example of a table tennis robot that performs competitively both in a simulation and on the real robot in a number of challenging scenarios.


Real-Time Spin Estimation of Ping-Pong Ball Using Its Natural Brand
A novel vision system that can provide both the position and the spin information of a flying ball in a real-time mode with high accuracy is proposed and experimental results show the effectiveness and precision of the proposed method.
A Table Tennis Robot System Using an Industrial KUKA Robot Arm
This work presents a novel table tennis robot system with high accuracy vision detection and fast robot reaction based on an industrial KUKA Agilus R900 sixx robot with 6 DOF, and tests both a curve fitting approach and an extended Kalman filter for predicting the ball’s trajectory.
Ball speed and spin estimation in table tennis using a racket-mounted inertial sensor
This contribution is the first attempt to estimate characteristics of rebounded balls with a single racket-mounted inertial sensor considering unknown initial conditions and constraints in table tennis.
Spin observation and trajectory prediction of a ping-pong ball
A way to observe and estimate ball's spin in real-time, and achieve an accurate prediction through Extended Kalman Filter (EKF) is proposed.
Estimating the spin of a table tennis ball using Inverse Compositional Image Alignment
A method for estimating the rotational velocity of a table tennis ball with Inverse Compositional Image Alignment, ICIA is proposed and an update rule for the motion parameters is derived.
Trajectory prediction of spinning ball for ping-pong player robot
An analytic flying model that can well represent the physical behavior is derived, where the ball's self-rotational velocity changes along with the flying velocity. Based on the least square method,
Model Based Motion State Estimation and Trajectory Prediction of Spinning Ball for Ping-Pong Robots using Expectation-Maximization Algorithm
This paper derives the Extended Continuous Motion Model (ECMM) by clustering the trajectories into multiple categories with a K-means algorithm and fitting them respectively using Fourier series and proposes a novel motion state estimation method using Expectation-Maximization (EM) algorithm, which in result contributes to an accurate trajectory prediction.
Optimal State Estimation of Spinning Ping-Pong Ball Using Continuous Motion Model
A model-based optimal algorithm for ball's motion state1 estimation is proposed using the initial trajectory acquired by a stereo vision system and it is proved that this optimization problem can be plotted as a convex optimization problem; thus, the globally optimal solution can be obtained.
Real-Time Deep Pose Estimation With Geodesic Loss for Image-to-Template Rigid Registration
The proposed deep learning-based methods that are trained to find the 3-D position of arbitrarily-oriented subjects or anatomy in a canonical space based on slices or volumes of medical images can dramatically enhance the performance of automatic imaging devices and image processing methods of the future.
3D Pose Regression Using Convolutional Neural Networks
It is argued that the 3D pose space is continuous and proposed to solve the pose estimation problem in a CNN regression framework with a suitable representation, data augmentation and loss function that captures the geometry of the pose space.