Rotation detection serves as a fundamental building block in many visual applications involving aerial image, scene text, and face etc. Differing from the dominant regression-based approaches for orientation estimation, this paper explores a relatively less-studied methodology based on classification. The hope is to inherently dismiss the boundary discontinuity issue as encountered by the regression-based detectors. We propose new techniques to push its frontier in two aspects: i) new encoding mechanism: the design of two Densely Coded Labels (DCL) for angle classification, to replace the Sparsely Coded Label (SCL) in existing classification-based detectors, leading to three times training speed increase as empirically observed across benchmarks, further with notable improvement in detection accuracy; ii) loss re-weighting: we propose Angle Distance and Aspect Ratio Sensitive Weighting (ADARSW), which improves the detection accuracy especially for square-like objects, by making DCL-based detectors sensitive to angular distance and object's aspect ratio. Extensive experiments and visual analysis on large-scale public datasets for aerial images i.e. DOTA, UCAS-AOD, HRSC2016, as well as scene text dataset ICDAR2015 and MLT, show the effectiveness of our approach. The source code is available at https://github.com/Thinklab-SJTU/DCL_RetinaNet_Tensorflow and is also integrated in our open source rotation detection benchmark: https://github.com/yangxue0827/RotationDetection.
Learning Polar Encodings for Arbitrary-Oriented Ship Detection in SAR Images
School of Electronic and Information Engineering, Beihang University（北京航空航天大学电子与信息工程学院）；Edinburgh Napier University（爱丁堡龙比亚大学）；Taibah University（塔伊巴大学）；University of Strathclyde（斯特拉斯克莱德大学）；University of Leicester（莱斯特大学）
常用的水平限界框（HBB）方法不能准确定位合成孔径雷达（SAR）图像中任意方位的细长舰船目标。因此，近年来，基于有向限界框（OBB）的方法逐渐受到研究者的关注。然而，最近提出的大多数基于深度学习的OBB检测方法都会遇到角度或关键点回归中的边界不连续问题。为了缓解这个问题，研究人员提出引入一些手动设置的参数或额外的网络分支来区分边界情况，这使得训练更加困难，导致性能下降。为了解决OBB回归中的边界不连续问题，本文提出了通过学习极坐标编码来检测SAR舰船。该编码方案使用一组从目标中心指向边界点的矢量来表示OBB。直接根据极坐标编码进行训练和推理，避免了边界不连续问题。此外，我们提出了一个Intersect over Union（IOU）加权回归损失，通过IOU度量进一步指导极性编码的训练，提高检测性能。在旋转SAR舰船检测数据集（RSSDD）上的实验结果表明，该方法的检测性能优于其他算法和OBB编码方案，证明了该方法的有效性。
ABFL: Angular Boundary Discontinuity Free Loss for Arbitrary Oriented Object Detection in Aerial Images
Technology and Engineering Center for Space Utilization, Chinese Academy of Sciences（中国科学院空间应用工程与技术中心）；University of Chinese Academy of Sciences（中国科学院大学）
Rethinking Boundary Discontinuity Problem for Oriented Object Detection
Hangzhou Dianzi University（杭州电子科技大学）；Institute of Computing Technology, Chinese Academy of Sciences（中国科学院计算技术研究所）；University of Chinese Academy of Sciences（中国科学院大学）；Lishui Institute of Hangzhou Dianzi University（杭州电子科技大学丽水研究院）
RSDet++: Point-based Modulated Loss for More Accurate Rotated Object Detection
Institute of Automation, Chinese Academy of Sciences（中国科学院自动化研究所）；University of Chinese Academy of Sciences（中国科学院大学）；Department of Computer Science and Engineering, Shanghai Jiao Tong University（上海交通大学计算机科学与工程系）；Inner Mongolia Key Laboratory of Molecular Biology on Featured Plants（内蒙古特色植物分子生物学重点实验室）
Learning Modulated Loss for Rotated Object Detection
University of Chinese Academy of Sciences（中国科学院大学）；Institute of Automation, Chinese Academy of Sciences（中国科学院自动化研究所）；Shanghai Jiao Tong University（上海交通大学）
