Learning Footstep Prediction from Motion Capture
نویسندگان
چکیده
Central pattern generated walking for bipedal robots has proven to be a versatile and easily implementable solution that is used by several robot soccer teams in the RoboCup Humanoid Soccer League with great success. However, the forward character of generating motor commands from an abstract, rhythmical pattern does not inherently provide the means for controlling the precise location of footsteps. For implementing a footstep planning gait control, we developed a step prediction model that estimates the location of the next footstep in Cartesian coordinates given the same inputs that control the central pattern generator. We used motion capture data recorded from walking robots to estimate the parameters of the prediction model and to verify the accuracy of the predicted footstep locations. We achieved a precision with a mean error of 1.3 cm.
منابع مشابه
Planning Biped Navigation Strategies in Complex Environments
We present an algorithm for planning goal-directed footstep navigation strategies for biped robots through obstacle-filled environments and uneven ground. Planning footsteps is more general than most existing navigation methods designed for wheeled robots, since the options of stepping over or upon obstacles are available. Given a height map of the terrain and a discrete set of possible footste...
متن کاملLearning Vector Quantization in Footstep Identification
This paper reports experiments on recognizing walkers from measurements with a pressure-sensitive floor, more specifically, a floor covered with EMFi material. A 100 square meter pressure-sensitive floor (EMFi floor) was recently installed in the Intelligent Systems Group’s research laboratory at the University of Oulu as part of a smart living room. The floor senses the changes in the pressure...
متن کاملReject-Optional LVQ-Based Two-Level Classifier to Improve Reliability in Footstep Identification
This paper reports experiments of recognizing walkers based on measurements with a pressure-sensitive EMFi-floor. Identification is based on a twolevel classifier system. The first level performs Learning Vector Quantization (LVQ) with a reject option to identify or to reject a single footstep. The second level classifies or rejects a sequence of three consecutive identified footsteps based on ...
متن کاملPrediction-driven Respiratory Motion Atlas Formation for 4D Image-guided Radiation Therapy in Lung
Respiratory motion challenges lung radiation therapy with uncertainties of the location of important anatomical structures in the thorax. To capture the trajectory of the motion, dense image matching methods and learning-based motion prediction methods have been commonly used. However, both methods have limitations. Serious motion artifacts in treatment-guidance images, such as streak artifacts...
متن کاملPrediction of Ground Reaction Forces and Moments via Supervised Learning Is Independent of Participant Sex, Height and Mass
Accurate multidimensional ground reaction forces and moments (GRF/Ms) can be predicted from marker-based motion capture using Partial Least Squares (PLS) supervised learning. In this study, the correlations between known and predicted GRF/Ms are compared depending on whether the PLS model is trained using the discrete inputs of sex, height and mass. All three variables were found to be accounte...
متن کامل