Search

In this paper we improve learning performance of a risk-aware robot facing navigation tasks by employing transfer learning; that is, we use information from a previously solved task to accelerate learning in a new task. To do so, we transfer risk-aware memoryless stochastic abstract policies into a new task. We show how to incorporate risk-awareness into robotic navigation tasks, in particular when tasks are modeled as stochastic shortest path problems. We then show how to use a modified policy iteration algorithm, called AbsProb-PI, to obtain risk-neutral and risk-prone memoryless stochastic abstract policies. Finally, we propose a method that combines abstract policies, and show how to use the combined policy in a new navigation task. Experiments validate our proposals and show that one can find effective abstract policies that can improve robot behavior in navigation problems.

Orthopedics is a natural and fitting target for the application of recent advances in imaging technology. In an effort to achieve accurate implant placement with minimal invasiveness, many have utilized robotic navigation systems such as the MAKO RIO®.

In this work we tackle the issue of visually recognising a place without any prior knowledge of its position, even in a world where the same place can look different or many places can look identical.

Self-localization is the basis of autonomous ability for mobile robotic navigation, motion planning. In order to meet the requirement of real-time performance of the game, an improved radial scan method to extract white lines of space is proposed in this paper, then use the information of white line to achieve self-localization for Soccer Robot, and the results of experiments show that the new method reduces not only the noise interference but also the amount of calculation, which could considerably enhance the real-time performance and accuracy of the system of Soccer Robot for MSRL.

Relational representations let sequential decision problems be described through objects and relations, resulting in more compact, expressive, and domain-independent representations that make it possible to find and generalize solutions much faster than propositional representations. In this paper we propose a modified policy iteration algorithm (AbsProb-PI) for the infinite-horizon discounted-reward criterion; the algorithm finds a memoryless probabilistic relational abstract policy that abstracts well the solution from source problems so that it can be applied in new, similar problems. Experiments in robotic navigation validate our proposals and show that we can find effective and efficient abstract policies, outperforming solutions by inductive approaches in the literature.

This chapter attempts to show how cognitive map models can be combined with robotic navigation strategies. A neural cognitive mapping strategy that is inspired by place cells but still abstract enough to be interpreted in a meaningful way is implemented in different experiments with both mobile robot and simulation experiments.

I will talk about recent and current work at Oxford on the automatic reconstruction of 3D information from 2D image sequences, and the applications of this technology to robotic navigation, augmented reality and the movie industry. The results of our work have been used on such movies as the “Lord of the Rings’” and “Harry Potter” series, and in 2002 we received an Emmy award for our contributions to television.