Publications | Max Planck Institute for Intelligent Systems

872 results (View BibTeX file of all listed publications)

2010

Stochastic Differential Dynamic Programming

Theodorou, E., Tassa, Y., Todorov, E.

In the proceedings of American Control Conference (ACC 2010) , 2010, clmc (article)

Abstract

We present a generalization of the classic Differential Dynamic Programming algorithm. We assume the existence of state- and control-dependent process noise, and proceed to derive the second-order expansion of the cost-to-go. Despite having quartic and cubic terms in the initial expression, we show that these vanish, leaving us with the same quadratic structure as standard DDP.

PDF [BibTex]

2010

am Theodorou, E., Tassa, Y., Todorov, E. Stochastic Differential Dynamic Programming In the proceedings of American Control Conference (ACC 2010) , 2010, clmc (article)

PDF [BibTex]

Learning Policy Improvements with Path Integrals

Theodorou, E. A., Buchli, J., Schaal, S.

In International Conference on Artificial Intelligence and Statistics (AISTATS 2010), 2010, clmc (inproceedings)

Abstract

With the goal to generate more scalable algo- rithms with higher efficiency and fewer open parameters, reinforcement learning (RL) has recently moved towards combining classi- cal techniques from optimal control and dy- namic programming with modern learning techniques from statistical estimation the- ory. In this vein, this paper suggests the framework of stochastic optimal control with path integrals to derive a novel approach to RL with parametrized policies. While solidly grounded in value function estimation and optimal control based on the stochastic Hamilton-Jacobi-Bellman (HJB) equations, policy improvements can be transformed into an approximation problem of a path inte- gral which has no open parameters other than the exploration noise. The resulting algorithm can be conceived of as model- based, semi-model-based, or even model free, depending on how the learning problem is structured. Our new algorithm demon- strates interesting similarities with previous RL research in the framework of proba- bility matching and provides intuition why the slightly heuristically motivated proba- bility matching approach can actually per- form well. Empirical evaluations demon- strate significant performance improvements over gradient-based policy learning and scal- ability to high-dimensional control problems. We believe that Policy Improvement with Path Integrals (PI2) offers currently one of the most efficient, numerically robust, and easy to implement algorithms for RL based on trajectory roll-outs.

PDF [BibTex]

am Theodorou, E. A., Buchli, J., Schaal, S. Learning Policy Improvements with Path Integrals In International Conference on Artificial Intelligence and Statistics (AISTATS 2010), 2010, clmc (inproceedings)

PDF [BibTex]

Learning optimal control solutions: a path integral approach

Theodorou, E., Schaal, S.

In Abstracts of Neural Control of Movement Conference (NCM 2010), Naples, Florida, 2010, 2010, clmc (inproceedings)

Abstract

Investigating principles of human motor control in the framework of optimal control has had a long tradition in neural control of movement, and has recently experienced a new surge of investigations. Ideally, optimal control problems are addresses as a reinforcement learning (RL) problem, which would allow to investigate both the process of acquiring an optimal control solution as well as the solution itself. Unfortunately, the applicability of RL to complex neural and biomechanics systems has been largely impossible so far due to the computational difficulties that arise in high dimensional continuous state-action spaces. As a way out, research has focussed on computing optimal control solutions based on iterative optimal control methods that are based on linear and quadratic approximations of dynamical models and cost functions. These methods require perfect knowledge of the dynamics and cost functions while they are based on gradient and Newton optimization schemes. Their applicability is also restricted to low dimensional problems due to problematic convergence in high dimensions. Moreover, the process of computing the optimal solution is removed from the learning process that might be plausible in biology. In this work, we present a new reinforcement learning method for learning optimal control solutions or motor control. This method, based on the framework of stochastic optimal control with path integrals, has a very solid theoretical foundation, while resulting in surprisingly simple learning algorithms. It is also possible to apply this approach without knowledge of the system model, and to use a wide variety of complex nonlinear cost functions for optimization. We illustrate the theoretical properties of this approach and its applicability to learning motor control tasks for reaching movements and locomotion studies. We discuss its applicability to learning desired trajectories, variable stiffness control (co-contraction), and parameterized control policies. We also investigate the applicability to signal dependent noise control systems. We believe that the suggested method offers one of the easiest to use approaches to learning optimal control suggested in the literature so far, which makes it ideally suited for computational investigations of biological motor control.

[BibTex]

am Theodorou, E., Schaal, S. Learning optimal control solutions: a path integral approach In Abstracts of Neural Control of Movement Conference (NCM 2010), Naples, Florida, 2010, 2010, clmc (inproceedings)

[BibTex]

Enhancing the performance of Bio-inspired adhesives

Chung, H., Glass, P., Sitti, M., Washburn, N. R.

In ABSTRACTS OF PAPERS OF THE AMERICAN CHEMICAL SOCIETY, 240, 2010 (inproceedings)

[BibTex]

pi Chung, H., Glass, P., Sitti, M., Washburn, N. R. Enhancing the performance of Bio-inspired adhesives In ABSTRACTS OF PAPERS OF THE AMERICAN CHEMICAL SOCIETY, 240, 2010 (inproceedings)

[BibTex]

Enhanced wet adhesion and shear of elastomeric micro-fiber arrays with mushroom tip geometry and a photopolymerized p (DMA-co-MEA) tip coating

Glass, P., Chung, H., Washburn, N. R., Sitti, M.

Langmuir, 26(22):17357-17362, American Chemical Society, 2010 (article)

Project Page [BibTex]

pi Glass, P., Chung, H., Washburn, N. R., Sitti, M. Enhanced wet adhesion and shear of elastomeric micro-fiber arrays with mushroom tip geometry and a photopolymerized p (DMA-co-MEA) tip coating Langmuir, 26(22):17357-17362, American Chemical Society, 2010 (article)

Project Page [BibTex]

Control performance simulation in the design of a flapping wing micro-aerial vehicle

Hines, L. L., Arabagi, V., Sitti, M.

In Intelligent Robots and Systems (IROS), 2010 IEEE/RSJ International Conference on, pages: 1090-1095, 2010 (inproceedings)

Project Page [BibTex]

pi Hines, L. L., Arabagi, V., Sitti, M. Control performance simulation in the design of a flapping wing micro-aerial vehicle In Intelligent Robots and Systems (IROS), 2010 IEEE/RSJ International Conference on, pages: 1090-1095, 2010 (inproceedings)

Project Page [BibTex]

Towards Semantic Scene Analysis with Time-of-flight Cameras

Holz, D., Schnabel, R., Droeschel, D., Stueckler, J., Behnke, S.

In RobuCup, 6556, pages: 121-132, Lecture Notes in Computer Science, Springer, 2010 (inproceedings)

link (url) [BibTex]

ev Holz, D., Schnabel, R., Droeschel, D., Stueckler, J., Behnke, S. Towards Semantic Scene Analysis with Time-of-flight Cameras In RobuCup, 6556, pages: 121-132, Lecture Notes in Computer Science, Springer, 2010 (inproceedings)

link (url) [BibTex]

On Parameter Learning in CRF-based Approaches to Object Class Image Segmentation

Nowozin, S., Gehler, P., Lampert, C.

In European Conference on Computer Vision, 2010 (inproceedings)

pdf supplementary [BibTex]

ps Nowozin, S., Gehler, P., Lampert, C. On Parameter Learning in CRF-based Approaches to Object Class Image Segmentation In European Conference on Computer Vision, 2010 (inproceedings)

pdf supplementary [BibTex]

Layered image motion with explicit occlusions, temporal consistency, and depth ordering

Sun, D., Sudderth, E., Black, M. J.

In Advances in Neural Information Processing Systems 23 (NIPS), pages: 2226-2234, MIT Press, 2010 (inproceedings)

Abstract

Layered models are a powerful way of describing natural scenes containing smooth surfaces that may overlap and occlude each other. For image motion estimation, such models have a long history but have not achieved the wide use or accuracy of non-layered methods. We present a new probabilistic model of optical flow in layers that addresses many of the shortcomings of previous approaches. In particular, we define a probabilistic graphical model that explicitly captures: 1) occlusions and disocclusions; 2) depth ordering of the layers; 3) temporal consistency of the layer segmentation. Additionally the optical flow in each layer is modeled by a combination of a parametric model and a smooth deviation based on an MRF with a robust spatial prior; the resulting model allows roughness in layers. Finally, a key contribution is the formulation of the layers using an image dependent hidden field prior based on recent models for static scene segmentation. The method achieves state-of-the-art results on the Middlebury benchmark and produces meaningful scene segmentations as well as detected occlusion regions.

main paper supplemental material paper and supplemental material in one pdf file Project Page [BibTex]

ps Sun, D., Sudderth, E., Black, M. J. Layered image motion with explicit occlusions, temporal consistency, and depth ordering In Advances in Neural Information Processing Systems 23 (NIPS), pages: 2226-2234, MIT Press, 2010 (inproceedings)

main paper supplemental material paper and supplemental material in one pdf file Project Page [BibTex]

Manifold Valued Statistics, Exact Principal Geodesic Analysis and the Effect of Linear Approximations

Stefan Sommer, Francois Lauze, Soren Hauberg, Mads Nielsen

In Computer Vision – ECCV 2010, 6316, pages: 43-56, (Editors: Daniilidis, Kostas and Maragos, Petros and Paragios, Nikos), Springer Berlin Heidelberg, 2010 (inproceedings)

Publishers site PDF [BibTex]

ps Stefan Sommer, Francois Lauze, Soren Hauberg, Mads Nielsen Manifold Valued Statistics, Exact Principal Geodesic Analysis and the Effect of Linear Approximations In Computer Vision – ECCV 2010, 6316, pages: 43-56, (Editors: Daniilidis, Kostas and Maragos, Petros and Paragios, Nikos), Springer Berlin Heidelberg, 2010 (inproceedings)

Publishers site PDF [BibTex]

GPU Accelerated Likelihoods for Stereo-Based Articulated Tracking

Rune Mollegaard Friborg, Soren Hauberg, Kenny Erleben

In The CVGPU workshop at European Conference on Computer Vision (ECCV) 2010, 2010 (inproceedings)

PDF [BibTex]

ps Rune Mollegaard Friborg, Soren Hauberg, Kenny Erleben GPU Accelerated Likelihoods for Stereo-Based Articulated Tracking In The CVGPU workshop at European Conference on Computer Vision (ECCV) 2010, 2010 (inproceedings)

PDF [BibTex]