interesting images for underwater research"
We have been conducting open water experiments with autonomous and semi-autonomous underwater robots for several years. While
the objective of the project has been to collect data that could be used by collaborators who study coral reef ecosystems, our work has dealt with autonomy, underwater human-robot interaction using computer vision, gait selection and learning, and behavior control. Most recently, we have been looking at multi-vehicle robotic
In this informal presentation I will discuss work in progress towards allowing
our robotic vehicle to automatically collect data for later use. In particular, we are
interested in a terse "Navigation Summary," that provides a synopsis of where the robot has been, and what is has seen. While this entails some human interaction, the core of the problem relates to the use of a model of novelty, specifically Bayesian
Surprise, to select novel images to summarize the experience of the robot. A key aspect of the
task is the use of a selection algorithm, akin to the classic "Secretaries Hiring
Problem," to select summary images.
The talk will entail a synopsis of the project and it's key themes, introduce
the "Navigation Summary" problem, and suggest some approaches regarding how it can be addresses.