Abstract
Object detection in remote sensing images is challenging due to the absence of visible features and variations in object orientation. Efficient detection of objects in such images can be achieved using rotated object detectors that utilize oriented bounding boxes. However, existing rotated object detectors often struggle to maintain high accuracy while processing high-resolution remote sensing images in real time. In this paper, we present RTMDet-R2, an improved real-time rotated object detector. RTMDet-R2 incorporates an enhanced path PAFPN to effectively fuse multi-level features and employs a task interaction decouple head to alleviate the imbalance between regression and classification tasks. To further enhance performance, we propose the ProbIoU-aware dynamic label assignment strategy, which enables efficient and accurate label assignment during the training. As a result, RTMDet-R2-m and RTMDet-R2-l achieve 79.10% and 79.46% mAP, respectively, on the DOTA 1.0 dataset using single-scale training and testing, outperforming the majority of other rotated object detectors. Moreover, RTMDet-R2-s and RTMDet-R2-t achieve 78.43% and 77.27% mAP, respectively, while achieving inference frame rates of 175 and 181 FPS at a resolution of 1024 × 1024 on an RTX 3090 GPU with TensorRT and FP16-precision. Furthermore, RTMDet-R2-t achieves 90.63/97.44% mAP on the HRSC2016 dataset. The code and models are available at https://github.com/Zeba-Xie/RTMDet-R2.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
We follow the latest metrics from the DOTA evaluation server, original voc format mAP is now mAP50.
References
Xie, X., Cheng, G.: Oriented R-CNN for object detection. In: ICCV, pp. 3520–3529 (2021)
Han, J., Ding, J., Xue, N., Xia, G.: ReDet: a rotation-equivariant detector for aerial object detection. In: CVPR, pp. 2786–2795 (2021)
Xue, Y., Qing, L., Junchi, Y.: R3det: Refined single-stage detector with feature refinement for rotating object. arXiv preprint arXiv:1908.05612 (2019)
Jiaming, H., Jian, D., Jie, L.: Align deep features for oriented object detection. IEEE Trans. Geosci. Aerial 60, 1–11 (2021)
Youtian, L., Pengming, F.: IENet: Interacting Embranchment One Stage Anchor Free Detector for Orientation Aerial Object Detection. arXiv preprint arXiv: 1912.00969 (2019)
Steven, L., Fabrizio, V., Kristian, K.: Dafne: A one-stage anchor-free deep model for oriented object detection. arXiv preprint arXiv:2109.06148 (2021)
Zhonghua, L., Biao, H., Zitong, W.: FCOSR: A simple anchor-free rotated detector for aerial object detection. arXiv preprint arXiv:2111.10780 (2021)
Wang, X., Wang, G., Dang, Q.: PP-YOLOE-R: An Efficient Anchor-Free Rotated Object Detector. arXiv preprint arXiv:2211.02386 (2022)
Lyu, C., Zhang, W., Huang, H.: RTMDet: An Empirical Study of Designing Real-Time Object Detectors. arXiv preprint arXiv:2212.07784 (2022)
Lin, T., Dollár, P., Girshick, R., He, K.: Feature pyramid networks for object detection. In: CVPR, pp. 2117–2125 (2017)
Liu, S., Qi, L., Qin, H.: Path aggregation network for instance segmentation. In: CVPR, pp. 8759–8768 (2018)
Chen, K., Cao, Y.: Feature pyramid grids. arXiv preprint arXiv:2004.03580 (2020)
Zheng G., Songtao L., Jian S.: YOLOX: Exceeding YOLO series in 2021. arXiv preprint arXiv:2107.08430 (2021)
Chengjian, F., Yujie, Z., Yu G.: TOOD: Task-aligned one-stage object detection. In: ICCV, pp. 3490–3499 (2021)
Shifeng, Z., Cheng, C., Yongqiang, Y.: Bridging the gap between anchor-based and anchor-free detection via adaptive training sample selection. In: CVPR, pp. 9759–9768 (2020)
Yue, W., Yinpeng, C., Lu, Y.: Rethinking classification and localization for object detection. In: CVPR, pp. 10186–10195 (2020)
Zhuang, J., Qin, Z., Yu, H., Chen, X.: Task-Specific Context Decoupling for Object Detection. arXiv preprint arXiv:2303.01047 (2023)
Jeffri, M.L., Luis, F.Z., Lucas, N.K., Claudio, J.: Gaussian bounding boxes and probabilistic intersection-over-union for object detection. arXiv preprintarXiv:2106.06072 (2021)
Xue, Y., Junchi, Y., Qi, M.: Rethinking rotated object detection with Gaussian Wasserste in distance loss. In: ICML, pp. 11830–11841 (2021)
Xue, Y., Yue, Z., Gefan, Z.: The kfiou loss for rotated object detection. arXiv preprint arXiv:2201.12558 (2022)
GuiSong, X., Xiang, B., Jian, D.: Dota: a large-scale dataset for object detection in aerial images. In: CVPR, pp. 3974–3983 (2018)
Zikun, L., Liu, Y., Lubin, W.: A high resolution optical satellite image dataset for ship recognition and some new baselines. In ICPRAM, pp. 324–331 (2017)
Zhou, Y., Yang, X., Zhang, G.: Mmrotate: a rotated object detection benchmark using pytorch. In: Proceedings of the 30th ACM International Conference on Multimedia, pp. 7331–7334 (2022)
Shaoqing R., Kaiming H., Ross G., Jian S.: Faster R-CNN: towards real-time object detection with region proposal networks. Adv. Neural Inf. Process. Syst. 28 (2015)
Kaiming, H., Georgia, G., Piotr, D.: Mask R-CNN. In: ICCV, pp. 2980–2988 (2017)
Joseph R., Ali F.: Yolov3: An incremental improvement. arXiv preprint arXiv:1804.02767 (2018)
ChienYao, W., Alexey, B.: Yolov7: trainable bag-of-freebies sets new state-of-the-art for real-time object detectors. In: CVPR, pp. 7464–7475 (2022)
Liu, W., Anguelov, D., Erhan, D.: SSD: single shot multibox detector. In: ECCV, pp. 21–37 (2016)
Cai, Z., Fan, Q., Feris, R.S., Vasconcelos, N.: A unified multi-scale deep convolutional neural network for fast object detection. In: ECCV, pp. 354–370 (2016)
Yu, Y., Da, F.: Phase-shifting coder: predicting accurate orientation in oriented object detection. In: CVPR, pp. 13354–13363 (2023)
Wang, J., Ding, J., Guo, H., Cheng, W., Pan, T.: Mask OBB: a semantic attention-based mask oriented bounding box representation for multi-category object detection in aerial images. Remote Sens. 11(24), 2930 (2019)
Li, W., Chen, Y., Hu, K., Zhu, J.: Oriented reppoints for aerial object detection. In: CVPR, pp. 1829–1838 (2022)
Guo, Z., Liu, C., Zhang, X., Jiao, J.: Beyond bounding-box: convex-hull feature adaptation for oriented and densely packed object detection. In: CVPR, pp. 8792–8801 (2021)
Yang, X., Yan, J.: Arbitrary-oriented object detection with circular smooth label. In: ECCV, pp. 677–694 (2020)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2024 The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.
About this paper
Cite this paper
Xiang, H. et al. (2024). RTMDet-R2: An Improved Real-Time Rotated Object Detector. In: Liu, Q., et al. Pattern Recognition and Computer Vision. PRCV 2023. Lecture Notes in Computer Science, vol 14436. Springer, Singapore. https://doi.org/10.1007/978-981-99-8555-5_28
Download citation
DOI: https://doi.org/10.1007/978-981-99-8555-5_28
Published:
Publisher Name: Springer, Singapore
Print ISBN: 978-981-99-8554-8
Online ISBN: 978-981-99-8555-5
eBook Packages: Computer ScienceComputer Science (R0)