New Polyhedral Study of Graph Multi-Separator Problem for Image Segmentation

This paper introduces a polyhedral study of the graph multi-separator problem, proposed as an alternative to the lifted multicut problem for image segmentation. The authors characterize facets of the multi-separator polytope induced by integer linear programming inequalities and explore stronger inequalities. They also establish a totally dual integral description for paths and relate the multi-separator polytope to the boolean quadric polytope and the lifted multicut polytope. AI

IMPACT This research could lead to improved methods for image segmentation by offering a new approach to graph partitioning problems.

RANK_REASON The cluster contains a single academic paper detailing a new polyhedral study of a graph problem. [lever_c_demoted from research: ic=1 ai=0.4]

Read on arXiv cs.LG →

• boolean quadric polytope
• graph multi-separator problem
• image segmentation
• integer linear programming
• Irmai et al.
• lifted multicut problem
• multi-separator polytope
• Paths

• paper
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →