Papers › Cascade RPN: Delving into High-Quality Region Proposal Network with Adaptive Convolution
Cascade RPN: Delving into High-Quality Region Proposal Network with Adaptive Convolution
Thang Vu, Hyunjun Jang, Trung X. Pham, Chang D. Yoo
This paper considers an architecture referred to as Cascade Region Proposal Network (Cascade RPN) for improving the region-proposal quality and detection performance by \textit{systematically} addressing the limitation of the conventional RPN that \textit{heuristically defines} the anchors and \textit{aligns} the features to the anchors. First, instead of using multiple anchors with predefined scales and aspect ratios, Cascade RPN relies on a \textit{single anchor} per location and performs multi-stage refinement. Each stage is progressively more stringent in defining positive samples by starting out with an anchor-free metric followed by anchor-based metrics in the ensuing stages. Second, to attain alignment between the features and the anchors throughout the stages, \textit{adaptive convolution} is proposed that takes the anchors in addition to the image features as its input and learns the sampled features guided by the anchors. A simple implementation of a two-stage Cascade RPN achieves AR 13.4 points higher than that of the conventional RPN, surpassing any existing region proposal methods. When adopting to Fast R-CNN and Faster R-CNN, Cascade RPN can improve the detection mAP by 3.1 and 3.5 points, respectively. The code is made publicly available at \url{https://github.com/thangvubk/Cascade-RPN.git}.
In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.
Code
Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.
Code Syntology ran Syntology
Not run by Syntology. Nothing on this page verifies that the listed code works.
Tasks
Results from the paper archive 2025-07-28
| Task | Dataset | Model | Metric | Value | Rank at snapshot | Leaderboard | Report |
|---|---|---|---|---|---|---|---|
| Object Detection | COCO test-dev | Faster R-CNN (Cascade RPN) | AP50 | 58.9 | #190 of 225 | Archive leaderboard | report |
| Object Detection | COCO test-dev | Faster R-CNN (Cascade RPN) | AP75 | 44.5 | #190 of 225 | Archive leaderboard | report |
| Object Detection | COCO test-dev | Faster R-CNN (Cascade RPN) | APL | 52.6 | #190 of 225 | Archive leaderboard | report |
| Object Detection | COCO test-dev | Faster R-CNN (Cascade RPN) | APM | 42.8 | #190 of 225 | Archive leaderboard | report |
| Object Detection | COCO test-dev | Faster R-CNN (Cascade RPN) | APS | 22.0 | #190 of 225 | Archive leaderboard | report |
| Object Detection | COCO test-dev | Faster R-CNN (Cascade RPN) | Hardware Burden | 5G | #190 of 225 | Archive leaderboard | report |
| Object Detection | COCO test-dev | Faster R-CNN (Cascade RPN) | box mAP | 40.6 | #190 of 225 | Archive leaderboard | report |
| Object Detection | COCO test-dev | Fast R-CNN (Cascade RPN) | AP50 | 59.4 | #198 of 225 | Archive leaderboard | report |
| Object Detection | COCO test-dev | Fast R-CNN (Cascade RPN) | AP75 | 43.8 | #198 of 225 | Archive leaderboard | report |
| Object Detection | COCO test-dev | Fast R-CNN (Cascade RPN) | APL | 51.6 | #198 of 225 | Archive leaderboard | report |
| Object Detection | COCO test-dev | Fast R-CNN (Cascade RPN) | APM | 42.4 | #198 of 225 | Archive leaderboard | report |
| Object Detection | COCO test-dev | Fast R-CNN (Cascade RPN) | APS | 22.1 | #198 of 225 | Archive leaderboard | report |
| Object Detection | COCO test-dev | Fast R-CNN (Cascade RPN) | Hardware Burden | 5G | #198 of 225 | Archive leaderboard | report |
| Object Detection | COCO test-dev | Fast R-CNN (Cascade RPN) | box mAP | 40.1 | #198 of 225 | Archive leaderboard | report |
Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.
Methods
Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections