Joint Graph Decomposition and Node Labeling by Local Search.

We state a combinatorial optimization problem whose feasible solutions define both a decomposition and a node labeling of a given graph. This problem offers a common mathematical abstraction of seemingly unrelated computer vision tasks, including instance-separating semantic segmentation, articulated human body pose estimation and multiple object tracking. Conceptually, the problem we propose generalizes the unconstrained integer quadratic program and the minimum cost lifted multicut problem, both of which are NP-hard. In order to find feasible solutions efficiently, we define a local search algorithm that converges monotonously to a local optimum, offering a feasible solution at any time. To demonstrate the effectiveness of this algorithm in solving computer vision tasks, we report running times and competitive solutions for two above-mentioned applications.