Research map / Detection and segmentation / CNN features
RF-Net: An End-To-End Image Matching Network Based on Receptive Field
Xuelun Shen, Cheng Wang, Xin Li, Zenglei Yu, Jonathan Li, Chenglu Wen, Ming Cheng, Zijian He
Check 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 |
|---|---|---|---|
| use RF-Net instead of L2-Net+ORB | local feature matching pipeline · HP-illum | average match score: 0.705 -> 0.783 | Improved |
| use RF-Net instead of L2-Net+ORB | local feature matching pipeline · HP-view | average match score: 0.673 -> 0.808 | Improved |
| use RF-Net instead of Hard-Net+SURF | local feature matching pipeline · EF | average match score: 0.334 -> 0.453 | Improved |
| replace LF-Det with RF-Det (ablation) | LF-Det+LF-Des pipeline · HP-illum | average match score: 0.617 -> 0.720 | Improved |
| replace LF-Det with RF-Det (ablation) | LF-Det+LF-Des pipeline · HP-view | average match score: 0.566 -> 0.665 | Improved |
| replace LF-Det with RF-Det (ablation) | LF-Det+LF-Des pipeline · EF | average match score: 0.251 -> 0.325 | Improved |
| replace LF-Des with RF-Des (ablation) | LF-Det+LF-Des pipeline · HP-illum | average match score: 0.617 -> 0.744 | Improved |
| replace LF-Des with RF-Des (ablation) | LF-Det+LF-Des pipeline · HP-view | average match score: 0.566 -> 0.714 | Improved |
| replace LF-Des with RF-Des (ablation) | LF-Det+LF-Des pipeline · EF | average match score: 0.251 -> 0.361 | Improved |
| add the neighbor mask to descriptor loss (ablation) | RF-Net · HP-illum | average match score: 0.734 -> 0.783 | Improved |
| add the neighbor mask to descriptor loss (ablation) | RF-Net · HP-view | average match score: 0.753 -> 0.808 | Improved |
| add the neighbor mask to descriptor loss (ablation) | RF-Net · EF | average match score: 0.423 -> 0.453 | Improved |
Later papers that compared with this method
2 later accepted papers report RFNet as a baseline in their results tables; 2 of them report a better number on most of the results they share with it.
- UniMRSeg: Unified Modality-Relax Segmentation via Hierarchical Self-Supervised Compensation (NeurIPS 2025 · better on 2 of 2 shared results)
- Hierarchy Parsing for Image Captioning (ICCV 2019 · better on 3 of 3 shared results)
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- FC-GNN: Recovering Reliable and Accurate Correspondences from Interferences (CVPR 2024 · similarity 0.87)
- Receptive Field Block Net for Accurate and Fast Object Detection (ECCV 2018 · similarity 0.87)
- Correspondence Networks With Adaptive Neighbourhood Consensus (CVPR 2020 · similarity 0.85)
- Deep Semantic Feature Matching (CVPR 2017 · similarity 0.85)
- NCNet: Neighbourhood Consensus Networks for Estimating Image Correspondences (TPAMI 2020 · similarity 0.85)
- Image Matching from Handcrafted to Deep Features: A Survey (IJCV 2020 · 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.
