A graph-based model of object recognition self-learning

In this paper, we study the object recognition self-learning for robots.In particular, we consider the self-learning during solution of typicaltasks. We propose a graph-based model for self-learning. This model isbased on the problem of monochromatic path for given set of weights.We prove that the problem is NP-complete. We consider an approachto solve the problem. This approach is based on an explicit reductionfrom the problem to the satisfiability problem.