用点标注训练旋转目标检测，通过空间布局提升密集场景精度
Point2RBox-v2: Rethinking Point-supervised Oriented Object Detection with Spatial Layout Among Instances
- 将物体视为高斯分布，用重叠损失学习实例上界
- 基于沃罗诺伊图分水岭，学习实例下界，提升定位精度
- 利用图像增强前后输出一致性，捕捉尺寸与旋转变化规律
随着对旋转目标检测（OOD）需求的快速增长，从点标注中弱监督学习的方法受到广泛关注。本文重新思考这一挑战性任务，提出Point2RBox-v2，核心包含三项原则：1）高斯重叠损失，将物体建模为二维高斯分布，最小化其重叠以学习每个实例的上界；2）沃罗诺伊分水岭损失，通过对沃罗诺伊剖分进行分水岭操作，学习每个实例的下界；3）一致性损失，通过输入图像及其增强视图的输出对比，学习尺寸与旋转的变化规律。结合边缘损失和复制粘贴等技巧，模型进一步优化。据我们所知，Point2RBox-v2是首个探索实例间空间布局以实现点标注旋转目标检测的方法。该方案简洁轻量，在密集场景中表现优异：在DOTA/HRSC/FAIR1M数据集上分别达到62.61%/86.15%/34.71%的mAP。代码已开源。
原文摘要 · Abstract (English)
With the rapidly increasing demand for oriented object detection (OOD), recent research involving weakly-supervised detectors for learning OOD from point annotations has gained great attention. In this paper, we rethink this challenging task setting with the layout among instances and present Point2RBox-v2. At the core are three principles: 1) Gaussian overlap loss. It learns an upper bound for each instance by treating objects as 2D Gaussian distributions and minimizing their overlap. 2) Voronoi watershed loss. It learns a lower bound for each instance through watershed on Voronoi tessellation. 3) Consistency loss. It learns the size/rotation variation between two output sets with respect to an input image and its augmented view. Supplemented by a few devised techniques, e.g. edge loss and copy-paste, the detector is further enhanced. To our best knowledge, Point2RBox-v2 is the first approach to explore the spatial layout among instances for learning point-supervised OOD. Our solution is elegant and lightweight, yet it is expected to give a competitive performance especially in densely packed scenes: 62.61%/86.15%/34.71% on DOTA/HRSC/FAIR1M. Code is available at https://github.com/VisionXLab/point2rbox-v2.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。