— It's all about the relationships we create with technology, how we view the artefacts we design, and what we learn about ourselves in the process.
— We understand the world and plan our day not as a process from perception to action, but the reversed, as anticipations of sensorconsequences of actions.
Predictive Sequence Learning is a method for robot learning from demonstration, able to anticipate future sensory-motor events in response to the recent interaction history.
A full-scale implementation in Java, underlying most scientific results referenced here, is available at bitbucket.org/interactionlab/psl.
Below is a summary of Erik Billing's PhD Thesis, titled Cognition Rehearsed - Recognition and Reproduction of Demonstrated Behavior, defended January 26, 2012.
The work presented in this dissertation investigates techniques for Learning from Demonstration (LFD). LFD is a well established approach to robot learning, where a teacher demonstrates a behavior to a robot pupil. The dissertation focuses on LFD where a human teacher demonstrates a behavior by controlling the robot via teleoperation. After demonstration, the robot should be able to reproduce the demonstrated behavior under varying conditions. In particular, the dissertation investigates techniques where previous behavioral knowledge is used as bias for generalization of demonstrations.
The primary contribution of this work is the development and evaluation of a semi-reactive approach to LFD called Predictive Sequence Learning (PSL). PSL has many interesting properties applied as a learning algorithm for robots. Few assumptions are introduced and little task-specific configuration is needed. PSL can be seen as a variable-order Markov model that progressively builds up the ability to predict or simulate future sensory-motor events, given a history of past events. The knowledge base generated during learning can be used to control the robot, such that the demonstrated behavior is reproduced. The same knowledge base can also be used to recognize an on-going behavior by comparing predicted sensor states with actual observations. Behavior recognition is an important part of LFD, both as a way to communicate with the human user and as a technique that allows the robot to use previous knowledge as parts of new, more complex, controllers.
In addition to the work on PSL, this dissertation provides a broad discussion on representation, recognition, and learning of robot behavior. LFD-related concepts such as demonstration, repetition, goal, and behavior are defined and analyzed, with focus on how bias is introduced by the use of behavior primitives. This analysis results in a formalism where LFD is described as transitions between information spaces. Assuming that the behavior recognition problem is partly solved, ways to deal with remaining ambiguities in the interpretation of a demonstration are proposed.
The evaluation of PSL shows that the algorithm can efficiently learn and reproduce simple behaviors. The algorithm is able to generalize to previously unseen situations while maintaining the reactive properties of the system. As the complexity of the demonstrated behavior increases, knowledge of one part of the behavior sometimes interferes with knowledge of another parts. As a result, different situations with similar sensory-motor interactions are sometimes confused and the robot fails to reproduce the behavior.
One way to handle these issues is to introduce a context layer that can support PSL by providing bias for predictions. Parts of the knowledge base that appear to fit the present context are highlighted, while other parts are inhibited. Which context should be active is continually re-evaluated using behavior recognition. This technique takes inspiration from several neurocomputational models that describe parts of the human brain as an hierarchical prediction system. With behavior recognition active, continually selecting the most suitable context for the present situation, the problem of knowledge interference is significantly reduced and the robot can successfully reproduce also more complex behaviors.
Erik Billing is an Associate Senior Lecturer in Informatics at the Interaction Lab, University of Skövde, Sweden. He holds a master in cognitive science and a PhD in computing science, both from Umeå University.
Erik devotes most of his time to research in the borderlands between artificial intelligence, robotics, and cognitive science. Among other things, Erik has explored how robots can learn from human demonstrations through anticipation of action and developed an algorithm for robot learning, called Predictive Sequence Learning. Examples and reference implementations is found here on cognitionreversed.com.