4.7 Article

Dual-Aligned Oriented Detector

出版社

IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/TGRS.2022.3149780

关键词

Proposals; Feature extraction; Object detection; Location awareness; Detectors; Remote sensing; Head; Localization-guided detection head (LDH); oriented object detection; oriented proposal network (OPN)

资金

  1. National Science Foundation of China [62136007, 41871322, 42130112]
  2. Shaanxi Science Foundation for Distinguished Young Scholars [2021JC-16]
  3. Open Research Fund of Key Laboratory of Space Utilization, Chinese Academy of Sciences [LSU-KFJJ-2021-06]

向作者/读者索取更多资源

This article presents a two-stage oriented object detection method, called dual-aligned oriented detector (DODet), to address the spatial and feature misalignment problems. DODet uses an oriented proposal network to generate high-quality proposals and a localization-guided detection head to alleviate the feature misalignment between classification and localization. Extensive evaluations on multiple benchmarks show consistent and substantial improvements compared to baseline methods.
In the past few years, object detection in remote sensing images has achieved remarkable progress. However, the detection of oriented and densely packed objects are still unsatisfactory due to the following spatial and feature misalignments. 1) Most two-stage oriented detectors only introduce an orientation regression branch in the detection head, while still leverage horizontal proposals for classification and regression. This inevitably results in the spatial misalignment problem between horizontal proposals and oriented objects. 2) The features used for classification are in fact extracted from the region proposals which have shifted to the final predictions via the regression branch. This leads to the feature misalignment problem between the classification and the localization tasks. In this article, we present a two-stage oriented object detection method, termed dual-aligned oriented detector (DODet), toward evading the aforementioned problems of spatial and feature misalignments. In DODet, the first stage is an oriented proposal network (OPN), which generates high-quality oriented proposals via a novel representation scheme of oriented objects. The second stage is a localization-guided detection head (LDH) that aims at alleviating the feature misalignment between classification and localization. Comprehensive and extensive evaluations on three benchmarks, including DIOR-R, DOTA, and HRSC2016, indicate that our method could obtain consistent and substantial gains compared with the baseline method. The source code is publicly available at https://github.com/yanqingyao1994/DODet.

作者

我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。

评论

主要评分

4.7
评分不足

次要评分

新颖性
-
重要性
-
科学严谨性
-
评价这篇论文

推荐

暂无数据
暂无数据