Publications
Below is a list of all relevant publications authored by Robotics Forum members.
Results
- Showing results for:
- Reset all filters
Search results
-
Conference paperKormushev P, Nenchev DN, Calinon S, et al., 2011,
Upper-body Kinesthetic Teaching of a Free-standing Humanoid Robot
, Pages: 3970-3975 -
Journal articleKormushev P, Calinon S, Caldwell DG, 2011,
Imitation Learning of Positional and Force Skills Demonstrated via Kinesthetic Teaching and Haptic Input
, Advanced Robotics, Vol: 25, Pages: 581-603 -
Conference paperKormushev P, Calinon S, Saegusa R, et al., 2010,
Learning the skill of archery by a humanoid robot iCub
, Pages: 417-423 -
Conference paperKormushev P, Calinon S, Caldwell DG, 2010,
Approaches for Learning Human-like Motor Skills which Require Variable Stiffness During Execution
-
Conference paperKormushev P, Calinon S, Caldwell DG, 2010,
Robot Motor Skill Coordination with EM-based Reinforcement Learning
, Pages: 3232-3237 -
Conference paperSato F, Nishii T, Takahashi J, et al., 2010,
Whiteboard Cleaning Task Realization with HOAP-2
, Pages: 426-429 -
Journal articleKormushev P, Nomoto K, Dong F, et al., 2009,
Eligibility Propagation to Speed up Time Hopping for Reinforcement Learning
, Journal of Advanced Computational Intelligence and Intelligent Informatics, Vol: 13, No. 6 -
Conference paperKormushev P, Dong F, Hirota K, 2009,
Probability redistribution using time hopping for reinforcement learning
-
Journal articleKormushev P, Nomoto K, Dong F, et al., 2008,
Time manipulation technique for speeding up reinforcement learning in simulations
, Cybernetics and Information Technologies, Vol: 8, Pages: 12-24, ISSN: 1311-9702A technique for speeding up reinforcement learning algorithms by usingtime manipulation is proposed. It is applicable to failure-avoidance controlproblems running in a computer simulation. Turning the time of the simulationbackwards on failure events is shown to speed up the learning by 260% andimprove the state space exploration by 12% on the cart-pole balancing task,compared to the conventional Q-learning and Actor-Critic algorithms.
-
Conference paperYamazaki Y, Dong F, Masuda Y, et al., 2007,
Intent expression using eye robot for mascot robot system
-
Conference paperYamazaki Y, Dong F, Masuda Y, et al., 2007,
Fuzzy inference based mentality estimation for eye robot agent
-
Journal articleAgre G, Kormushev P, Dilov I, 2006,
INFRAWEBS Axiom Editor - A graphical ontology-driven tool for creating complex logical expressions
, International Journal of Information Theories and Applications, Vol: 13, Pages: 169-178 -
Conference paperAgre G, Kormushev P, Dilov I, 2005,
INFRAWEBS Capability Editor - A graphical ontology-driven tool for creating capabilities of Semantic Web Services
, Pages: 228-228 -
Journal articleVrielink TJCO, Pang YW, Zhao M, et al.,
Surgical task-space optimisation of the CYCLOPS robotic system
The CYCLOPS is a cable-driven parallel mechanism used for minimally invasiveapplications, with the ability to be customised to different surgical needs;allowing it to be made procedure- and patient-specific. For adequateoptimisation, however, appropriate data on clinical constraints and task-spaceis required. Whereas the former can be provided through preoperative planningand imaging, the latter remains a problem, primarily for highly dexterous MISsystems. The current work focuses on the development of a task-spaceoptimisation method for the CYCLOPS system and the development of a datacollection method in a simulation environment for minimally invasivetask-spaces. The same data collection method can be used for the development ofother minimally invasive platforms. A case-study is used to illustrate thedeveloped method for Endoscopic Submucosal Dissection (ESD). This paper showsthat using this method, the system can be succesfully optimised for thisapplication.
-
Journal articleChappell D, Wang K, Kormushev P,
Asynchronous Real-Time Optimization of Footstep Placement and Timing in Bipedal Walking Robots
Online footstep planning is essential for bipedal walking robots to be ableto walk in the presence of disturbances. Until recently this has been achievedby only optimizing the placement of the footstep, keeping the duration of thestep constant. In this paper we introduce a footstep planner capable ofoptimizing footstep placement and timing in real-time by asynchronouslycombining two optimizers, which we refer to as asynchronous real-timeoptimization (ARTO). The first optimizer which runs at approximately 25 Hz,utilizes a fourth-order Runge-Kutta (RK4) method to accurately approximate thedynamics of the linear inverted pendulum (LIP) model for bipedal walking, thenuses non-linear optimization to find optimal footsteps and duration at a lowerfrequency. The second optimizer that runs at approximately 250 Hz, usesanalytical gradients derived from the full dynamics of the LIP model andconstraint penalty terms to perform gradient descent, which finds approximatelyoptimal footstep placement and timing at a higher frequency. By combining thetwo optimizers asynchronously, ARTO has the benefits of fast reactions todisturbances from the gradient descent optimizer, accurate solutions that avoidlocal optima from the RK4 optimizer, and increases the probability that afeasible solution will be found from the two optimizers. Experimentally, weshow that ARTO is able to recover from considerably larger pushes and producesfeasible solutions to larger reference velocity changes than a standardfootstep location optimizer, and outperforms using just the RK4 optimizeralone.
-
Journal articleWong MZ, Guillard B, Murai R, et al.,
AnalogNet: Convolutional Neural Network Inference on Analog Focal Plane Sensor Processors
We present a high-speed, energy-efficient Convolutional Neural Network (CNN)architecture utilising the capabilities of a unique class of devices known asanalog Focal Plane Sensor Processors (FPSP), in which the sensor and theprocessor are embedded together on the same silicon chip. Unlike traditionalvision systems, where the sensor array sends collected data to a separateprocessor for processing, FPSPs allow data to be processed on the imagingdevice itself. This unique architecture enables ultra-fast image processing andhigh energy efficiency, at the expense of limited processing resources andapproximate computations. In this work, we show how to convert standard CNNs toFPSP code, and demonstrate a method of training networks to increase theirrobustness to analog computation errors. Our proposed architecture, coinedAnalogNet, reaches a testing accuracy of 96.9% on the MNIST handwritten digitsrecognition task, at a speed of 2260 FPS, for a cost of 0.7 mJ per frame.
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.