paper-with-me

홈 › Papers

A Continuous Max-Flow Approach to Multi-Labeling Problems under Arbitrary Region Regularization

2014-05-05 · John S. H. Baxter, Martin Rajchl, Jing Yuan, Terry M. Peters

The incorporation of region regularization into max-flow segmentation has traditionally focused on ordering and part-whole relationships. A side effect of the development of such models is that it constrained regularization only to those cases, rather than allowing for arbitrary region regularization. Directed Acyclic Graphical Max-Flow (DAGMF) segmentation overcomes these limitations by allowing for the algorithm designer to specify an arbitrary directed acyclic graph to structure a max-flow segmentation. This allows for individual 'parts' to be a member of multiple distinct 'wholes.'

📄 PDF Abstract BibTeX arXiv:1405.0892

Code (0)

등록된 구현이 없습니다.

Tasks

Segmentation

Similar Papers 제목 키워드 기반

Shape Complexes in Continuous Max-Flow Hierarchical Multi-Labeling Problems

2015-10-15 · John S. H. Baxter, Jing Yuan, Terry M. Peters

Although topological considerations amongst multiple labels have been previously investigated in the context of continuous max-flow image segmentation, similar investigations have yet to be made about shape consideration…

Image SegmentationSegmentationSemantic Segmentation

WeaNF: Weak Supervision with Normalizing Flows

2022-04-28 · Andreas Stephan, Benjamin Roth

A popular approach to decrease the need for costly manual annotation of large data sets is weak supervision, which introduces problems of noisy labels, coverage and bias. Methods for overcoming these problems have either…

WeaNF: Weak Supervision with Normalizing Flows

2021-11-16 · ACL ARR November 2021 11 · Anonymous

A popular approach to decrease the need for costly manual annotation of large data sets is weak supervision, which introduces problems of noisy labels, coverage and bias. Methods for overcoming these problems have either…

WeaNF”:" Weak Supervision with Normalizing Flows

2022-05-01 · RepL4NLP (ACL) 2022 5 · Andreas Stephan, Benjamin Roth

A popular approach to decrease the need for costly manual annotation of large data sets is weak supervision, which introduces problems of noisy labels, coverage and bias. Methods for overcoming these problems have either…

Actor-Action Semantic Segmentation with Grouping Process Models

2015-12-30 · CVPR 2016 6 · Chenliang Xu, Jason J. Corso

Actor-action semantic segmentation made an important step toward advanced video understanding problems: what action is happening; who is performing the action; and where is the action in space-time. Current models for th…

Semantic SegmentationVideo Understanding