Bootstrap Learning and Visual Processing Management on Mobile Robots
Figure 6
Example query: “where are the blue circles?” LL-POMDP reward specification results in early termination when negative evidence is found. Belief propagation provides reliability.
