AI Paper Atlas

Research map / Detection and segmentation / Weakly supervised detection and segmentation

Recurrent Transformer Networks for Semantic Correspondence

Seungryong Kim, Stephen Lin, SANG RYUL JEON, Dongbo Min, Kwanghoon Sohn

NeurIPS 2018 · Detection and segmentation › Weakly supervised detection and segmentation · approach: Weakly supervised semantic segmentation · 49 citations (OpenAlex)

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 changedWhereMeasuredEffect
Use recurrent geometric matching with ResNet features instead of GMat. w/Inl.Semantic correspondence · TSS averagematching accuracy: 0.737 -> 0.772Improved
Use RTNs with ResNet instead of the strongest listed baselineSemantic correspondence · PF-WILLOWPCK α=0.05: 0.386 -> 0.413Improved
Use RTNs with ResNet instead of the strongest listed baselineSemantic correspondence · PF-WILLOWPCK α=0.1: 0.704 -> 0.719Improved
Use RTNs with ResNet instead of the strongest listed baselineSemantic correspondence · PF-PASCALPCK α=0.05: 0.490 -> 0.552Improved
Use RTNs with ResNet instead of the strongest listed baselineSemantic correspondence · PF-PASCALPCK α=0.1: 0.748 -> 0.759Improved
Use RTNs with ResNet instead of the strongest listed baselineSemantic correspondence · PF-PASCALPCK α=0.15: 0.840 -> 0.852Improved
Use 3–5 recurrent iterations (ablation)RTNs w/ResNet · TSSmatching accuracyImproved
Enlarge the local search window up to 9×9 (ablation)RTNs · TSSmatching accuracyImproved
Enlarge the local search window beyond 9×9 (ablation)RTNs · TSSmatching accuracyHurt
Use ResNet features instead of VGGNet or CAT-FCSS features (ablation)RTNs · TSSmatching accuracyImproved
Jointly learn the feature extractor with the geometric matching network instead of using a pretrained backbone (ablation)RTNs · TSSmatching accuracyImproved

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy