Research map / Detection and segmentation / Weakly supervised detection and segmentation
FCSS: Fully Convolutional Self-Similarity for Dense Semantic Correspondence
Seungryong Kim, Dongbo Min, Bumsub Ham, Sangryul Jeon, Stephen Lin, Kwanghoon Sohn
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 |
|---|---|---|---|
| Replace VGG descriptors with FCSS | VGG-Net descriptor with fixed SIFT Flow optimization · Taniai average | matching accuracy: 0.535 -> 0.660 | Improved |
| Use Proposal Flow optimization instead of SIFT Flow | FCSS · Proposal Flow benchmark, α=0.1 | PCK: 0.532 -> 0.584 | Improved |
| Use Proposal Flow optimization instead of SIFT Flow | FCSS · PASCAL-VOC part, PCK α=0.1 | PCK: 0.47 -> 0.46 | No clear effect |
| Use learned sampling patterns rather than random sampling patterns (ablation) | VGG with single-scale CSS · Taniai average | matching accuracy: 0.551 -> 0.573 | Improved |
| Add learned single-scale CSS (ablation) | VGG-Net descriptor · Taniai average | matching accuracy: 0.535 -> 0.573 | Improved |
| Add multi-scale CSS instead of single-scale CSS (ablation) | VGG-Net descriptor · Taniai average | matching accuracy: 0.573 -> 0.610 | Improved |
| Add multi-scale CSS instead of single-scale CSS (ablation) | VGG-Net descriptor · Proposal Flow benchmark, α=0.1 | PCK: 0.422 -> 0.514 | Improved |
Closest papers by meaning
- Learning Contrastive Representation for Semantic Correspondence (IJCV 2022 · similarity 0.89)
- Semantic-Aware Fine-Grained Correspondence (ECCV 2022 · similarity 0.88)
- Learning Semantic Correspondence Exploiting an Object-Level Prior (TPAMI 2020 · similarity 0.88)
- Deep Semantic Feature Matching (CVPR 2017 · similarity 0.88)
- Weakly Supervised Manifold Learning for Dense Semantic Object Correspondence (ICCV 2017 · similarity 0.87)
- Semantic Correspondence: Unified Benchmarking and a Strong Baseline (TPAMI 2025 · similarity 0.87)
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.
