Research map / Detection and segmentation / CNN features
Correspondence Networks With Adaptive Neighbourhood Consensus
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| use ANC-Net instead of HPF ResNet-101-FCN | PF-PASCAL · PF-PASCAL | PCK@0.1: 88.3 -> 88.7 | Helped |
| use ANC-Net instead of DCCNet | CUB · CUB | PCK@0.1: 66.1 -> 74.1 | Helped |
| use ANC-Net instead of HPF ResNet-101 | Spair-71k · Spair-71k | PCK@0.1: 28.2 -> 30.1 | Helped |
| replace isotropic 4D kernels with adaptive non-isotropic ANC kernels (ablation) | ANC-Net · PF-PASCAL | PCK@0.1: 84.1 -> 86.1 | Helped |
| add multi-scale self-similarity (ablation) | ANC-Net · PF-PASCAL | PCK@0.1: 84.3 -> 86.1 | Helped |
| add orthogonal loss (ablation) | ANC-Net · PF-PASCAL | PCK@0.1: 85.9 -> 86.1 | Helped |
| use ANC-d rather than ANC-b (ablation) | ANC-Net · PF-PASCAL | PCK@0.1: 82.7 -> 86.1 | Helped |
| use ANC-d rather than ANC-c (ablation) | ANC-Net · PF-PASCAL | PCK@0.1: 83.8 -> 86.1 | Helped |
| replace ResNet-101 backbone with ResNeXt-101 | ANC-Net · PF-PASCAL | PCK@0.1: 86.1 -> 88.7 | Helped |
| replace ResNet-101 backbone with ResNeXt-101 | ANC-Net · CUB | PCK@0.1: 72.4 -> 74.1 | Helped |
| replace ResNet-101 backbone with ResNeXt-101 | ANC-Net · Spair-71k | PCK@0.1: 28.7 -> 30.1 | Helped |
Closest papers by meaning
- Neighbourhood Consensus Networks (NeurIPS 2018 · similarity 0.92)
- PatchMatch-Based Neighborhood Consensus for Semantic Correspondence (CVPR 2021 · similarity 0.89)
- Dual-Resolution Correspondence Networks (NeurIPS 2020 · similarity 0.88)
- NM-Net: Mining Reliable Neighbors for Robust Feature Correspondences (CVPR 2019 · similarity 0.87)
- Spotting the Unseen: Reciprocal Consensus Network Guided by Visual Archetypes (AAAI 2024 · similarity 0.86)
- ETO:Efficient Transformer-based Local Feature Matching by Organizing Multiple Homography Hypotheses (NeurIPS 2024 · similarity 0.85)
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.