Cascade RPN: Delving into High-Quality Region Proposal Network with Adaptive Convolution

By Thang Vu, Hyunjun Jang, Trung X. Pham, Chang-Dong Yoo. 2019

209 citations recorded by the source corpus.

This paper considers an architecture referred to as Cascade Region Proposal Network (Cascade RPN) for improving the region-proposal quality and detection performance by systematically addressing the limitation of the conventional RPN that heuristically defines the anchors and aligns the features to the anchors. First, instead of using multiple anchors with predefined scales and aspect ratios, Cascade RPN relies on a 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, 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 13.4 point AR 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 will be made publicly available at https://github.com/thangvubk/Cascade-RPN.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Vision and Pattern Recognition

Citations in the corpus

Learning Open-World Object Proposals without Learning to Classify (2021, methodology): “Researchers have studied many ways to improve the localization quality in object detections by learning centerness [14], iterative proposal refinement [5], or box/mask IoU prediction [15], [16], [36].”
Probabilistic two-stage detection (2021, methodology): “The models of GA RPN and Cascade RPN are from mmdetection (Chen et al., 2019b) test set.”
Learning to Match Anchors for Visual Object Detection (2021, background): “Cascade RPN [22] improved region-proposal quality and detection performance by addressing the limitation of the conventional RPN which heuristically deﬁnes anchors and aligns features to anchors.”
Ocean: Object-aware Anchor-free Tracking (2020, background): “The alignment between visual features and reference ROIs (Regions of Interests) is vital for localization tasks, such as detection and tracking [40].”
Anchor-Free Oriented Proposal Generator for Object Detection (2021, unknown): “The comparison of the coarse oriented boxes with the horizontal proposals generated by RPN is made in Section IV.”