CenterRCNN: Two-Stage Anchor-Free Object Detection Using Center Keypoint-Based Region Proposal Network

doi:10.1007/s12204-023-2667-y

Abstract

Abstract: The classic two-stage object detection algorithms such as faster regions with convolutional neural network features (Faster RCNN) suffer from low speed and anchor hyper-parameter sensitive problems caused by dense anchor mechanism in region proposal network (RPN). Recently, the anchor-free method CenterNet shows the effectiveness of perceiving and classifying object by its center. However, the severe coincidence false positive problem between confusing categories caused by the multiple binary classifiers makes it still insufficient in accuracy. We introduce a two-stage network CenterRCNN to take advantage of both and overcome their shortcomings. CenterRPN is proposed as the first stage to give proposals that incorporate the center keypoint idea into RPN to perceive foreground objects, replacing dense anchor-based RPN. Then the proposals are classified by the multi-classifier of RCNN header that focuses more on the difference between confusing categories and only outputs the maximum probability one of them. To sum up, CenterRPN can eliminate the drawbacks of dense anchor based RPN in Faster RCNN, and multi-classifier’s classification ability is better than that of multiple binary classifiers in CenterNet. The experiment demonstrates that CenterRCNN outperforms both basic algorithms in the accuracy, and the speed is improved as compared with Faster RCNN.

Key words: anchor-free detection, CenterRPN, multi-classifier

摘要： 经典的两阶段目标检测算法，例如快速区域卷积神经网络（Faster RCNN），由于区域提议网络（RPN）中密集锚框机制引起的低速度和锚框超参数敏感性问题而受到影响。最近，无锚框方法CenterNet通过其中心感知和分类对象展示有效性。然而，由于多个二进制分类器引起的混淆类之间的假阳问题使CenterNet仍然不够准确。我们引入了一个两阶段网络CenterRCNN来利用两者的优势并克服缺点。我们提出CenterRPN作为第一阶段，以将中心关键点思想纳入RPN中感知前景对象，并替换基于密集锚框的RPN。然后，候选框通过RCNN检测头的多分类器进行分类，该分类器更加关注混淆类之间的差异，并仅输出其中最大概率的一个。总之，CenterRPN可以消除Faster RCNN中基于密集锚框的RPN的缺点，并且多分类器的分类能力优于CenterNet中的多个二进制分类器。实验表明：CenterRCNN准确性优于两种基本算法，并且与Faster RCNN相比速度有所提高。

关键词: 无锚框检测，CenterRPN，多分类器

CLC Number:

TP391

LIU Chen, LI Wenfa, XU Yunwen, LI Dewei. CenterRCNN: Two-Stage Anchor-Free Object Detection Using Center Keypoint-Based Region Proposal Network[J]. J Shanghai Jiaotong Univ Sci, 2025, 30(5): 1028-1036.

References

[1] REN S Q, HE K M, GIRSHICK R, et al. Faster R-CNN: Towards real-time object detection with region proposal networks [J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2017, 39(6): 1137-1149.

[2] KISANTAL M, WOJNA Z, MURAWSKI J, et al. Augmentation for small object detection [DB/OL]. (2019-02-19). https://arxiv.org/abs/1902.07296

[3] LIN T Y, DOLLÁR P, GIRSHICK R, et al. Feature pyramid networks for object detection [C]//2017 IEEE Conference on Computer Vision and Pattern Recognition. Honolulu: IEEE, 2017: 936-944.

[4] CAI Z W, VASCONCELOS N. Cascade R-CNN: Delving into high quality object detection [C]//2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Salt Lake City: IEEE, 2018: 6154-6162.

[5] LIN T Y, GOYAL P, GIRSHICK R, et al. Focal loss for dense object detection [C]//2017 IEEE International Conference on Computer Vision. Venice: IEEE, 2017: 2999-3007.

[6] FU C Y, LIU W, RANGA A, et al. DSSD: Deconvolutional single shot detector [DB/OL]. (2017-01-23) https://arxiv.org/abs/1701.06659

[7] TAN M X, PANG R M, LE Q V. EfficientDet: scalable and efficient object detection [C]//2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Seattle: IEEE, 2020: 10778-10787.

[8] TAN M X, LE Q V. EfficientNet: Rethinking model scaling for convolutional neural networks [C]//International Conference on Machine Learning. Long Beach: IMLS, 2019: 6105-6114.

[9] LIU Z, LIN Y T, CAO Y, et al. Swin transformer: Hierarchical vision transformer using shifted windows [C]//2021 IEEE/CVF International Conference on Computer Vision. Montreal: IEEE, 2021: 9992-10002.

[10] LAW H, DENG J. CornerNet: detecting objects as paired keypoints[M]//European conference on computer vision. Cham: Springer, 2018: 765-781.

[11] ZHOU X Y, WANG D Q, KRÄHENBÜHL P. Objects as points [DB/OL]. (2019-04-16). https://arxiv.org/abs/1904.07850

[12] WANG J Q, CHEN K, YANG S, et al. Region proposal by guided anchoring [C]//2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Long Beach: IEEE, 2019: 2960-2969.

[13] EVERINGHAM M, WINN J. The PASCAL Visual Object Classes Challenge 2012 (VOC2012) development kit [EB/OL]. (2012-05-18). http://host.robots.ox.ac.uk/pascal/VOC/voc2012/devkit_doc.pdf

[14] LIN T Y, MAIRE M, BELONGIE S, et al. Microsoft COCO: Common objects in context[M]//European conference on computer vision. Cham: Springer, 2014: 740-755.

[15] LIU W, ANGUELOV D, ERHAN D, et al. SSD: single shot MultiBox detector[M]//European conference on computer vision. Cham: Springer, 2016: 21-37.

[16] HE K M, ZHANG X Y, REN S Q, et al. Deep residual learning for image recognition [C]//2016 IEEE Conference on Computer Vision and Pattern Recognition. Las Vegas: IEEE, 2016: 770-778.

[17] DAI J F, LI Y, HE K M, et al. R-FCN: Object detection via region-based fully convolutional networks [C]// 30th Conference on Neural Information Processing Systems. Barcelona: NIPS, 2016: 379-387.