Research map / Label-efficient and robust learning / Recognition and long-tail
An Inference Algorithm for Multi-Label MRF-MAP Problems with Clique Size 100
Ishant Shanu, Siddhant Bharti, Chetan Arora, S. N. Maheshwari
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| apply MLHybrid with small cliques on top of Deeplabv3+ output | Deeplabv3+ · Pascal VOC (noisy) | pixel accuracy: 82.79 -> 84.07 | Improved |
| increase clique size from small (60-80) to big (300-400) (ablation) | MLHybrid · Pascal VOC (noisy) | pixel accuracy: 84.07 -> 85.11 | Improved |
| increase clique size from small to big (ablation) | MLHybrid · Pascal VOC (noisy) | mean IoU: 0.566 -> 0.579 | Improved |
| apply MLHybrid with big cliques on top of Deeplabv3+ | Deeplabv3+ · Pascal VOC (noisy) | mean IoU: 0.544 -> 0.579 | Improved |
| use non-decomposable concave-of-cardinality potential instead of decomposable ABS potential (ablation) | MLHybrid · Pascal VOC (noisy) | segmentation quality and run time | Improved |
| fine-tune hyper parameters instead of standard fixed ones (ablation) | MLHybrid · Pascal VOC (noisy) | segmentation quality (IoU) | Improved |
| bypass invalid-state costs (MLHybrid) instead of large penalty L in SOS-MNP (ablation) | SOS-MNP · Stereo (sawtooth, 16 labels) | convergence time: approximately 50 minutes without convergence -> around 28 seconds | Improved |
Closest papers by meaning
- Min Norm Point Algorithm for Higher Order MRF-MAP Inference (CVPR 2016 · similarity 0.92)
- Inference in Higher Order MRF-MAP Problems With Small and Large Cliques (CVPR 2018 · similarity 0.91)
- Complexity of Discrete Energy Minimization Problems (ECCV 2016 · similarity 0.87)
- Sublabel-Accurate Convex Relaxation of Vectorial Multilabel Energies (ECCV 2016 · similarity 0.86)
- Automatically Selecting Inference Algorithms for Discrete Energy Minimisation (ECCV 2016 · similarity 0.86)
- MRF Optimization with Separable Convex Prior on Partially Ordered Labels (ECCV 2018 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
