Publications
Results
- Showing results for:
- Reset all filters
Search results
-
Conference paperKryczka P, Kormushev P, Tsagarakis N, et al., 2015,
Online Regeneration of Bipedal Walking Gait Optimizing Footstep Placement and Timing
-
Conference paperKormushev P, Demiris Y, Caldwell DG, 2015,
Kinematic-free Position Control of a 2-DOF Planar Robot Arm
-
Journal articleCalandra R, Seyfarth A, Peters J, et al., 2015,
Bayesian Optimization for Learning Gaits under Uncertainty
, Annals in Mathematics and Artificial Intelligence, Vol: 76, Pages: 5-23, ISSN: 1012-2443Designing gaits and corresponding control policies is a key challenge in robot locomotion. Even with a viable controller parametrization, finding near-optimal parameters can be daunting. Typically, this kind of parameter optimization requires specific expert knowledge and extensive robot experiments. Automatic black-box gait optimization methods greatly reduce the need for human expertise and time-consuming design processes. Many different approaches for automatic gait optimization have been suggested to date. However, no extensive comparison among them has yet been performed. In this article, we thoroughly discuss multiple automatic optimization methods in the context of gait optimization. We extensively evaluate Bayesian optimization, a model-based approach to black-box optimization under uncertainty, on both simulated problems and real robots. This evaluation demonstrates that Bayesian optimization is particularly suited for robotic applications, where it is crucial to find a good set of gait parameters in a small number of experiments.
-
Conference paperDeisenroth MP, Ng JW, 2015,
Distributed Gaussian Processes
, 2015 International Conference on Machine Learning (ICML), Publisher: Journal of Machine Learning ResearchTo scale Gaussian processes (GPs) to large datasets we introduce the robust Bayesian CommitteeMachine (rBCM), a practical and scalableproduct-of-experts model for large-scaledistributed GP regression. Unlike state-of-theartsparse GP approximations, the rBCM is conceptuallysimple and does not rely on inducingor variational parameters. The key idea is torecursively distribute computations to independentcomputational units and, subsequently, recombinethem to form an overall result. Efficientclosed-form inference allows for straightforwardparallelisation and distributed computations witha small memory footprint. The rBCM is independentof the computational graph and canbe used on heterogeneous computing infrastructures,ranging from laptops to clusters. With sufficientcomputing resources our distributed GPmodel can handle arbitrarily large data sets.
-
Journal articleCarrera A, Palomeras N, Hurtós N, et al., 2015,
Cognitive System for Autonomous Underwater Intervention
, Pattern Recognition Letters, ISSN: 0167-8655 -
Conference paperWahlstrom N, Schon TB, Deisenroth MP, 2015,
Learning Deep Dynamical Models From Image Pixels
, 17th IFAC Symposium on System Identification, SYSID 2015 -
Conference paperKormushev P, Demiris Y, Caldwell DG, 2015,
Encoderless Position Control of a Two-Link Robot Manipulator
-
Conference paperAhmadzadeh SR, Paikan A, Mastrogiovanni F, et al., 2015,
Learning Symbolic Representations of Actions from Human Demonstrations
-
Conference paperJamali N, Kormushev P, Carrera A, et al., 2015,
Underwater Robot-Object Contact Perception using Machine Learning on Force/Torque Sensor Feedback
-
Conference paperCarrera A, Palomeras N, Hurtos N, et al., 2015,
Learning multiple strategies to perform a valve turning with underwater currents using an I-AUV
This data is extracted from the Web of Science and reproduced under a licence from Thomson Reuters. You may not copy or re-distribute this data in whole or in part without the written consent of the Science business of Thomson Reuters.