.png){kind=icon}
返回
Fast arbitrary-oriented object detection for remote sensing images
.png){kind=icon}
OA
AI
DOI：10.1080/22797254.2024.2431006.png)
.png){kind=icon}
原文求助.png){kind=icon}
分享.png){kind=icon}
收藏
摘要
En 中文
In the filed of remote sensing, arbitrary-oriented object detection methods has gained great attention, benefiting from the accurate detection ability of dense objects. However, the existing methods, which are designed based on ResNet, are not fast enough for real-time application. To solve this problem, our paper proposes a new fast arbitrary-oriented object detection methods based on YOLOX. First, a new head for rotational box prediction is proposed, in which a new branch is designed to extract the angle information through weighted averaging from different angles. Then, a new loss function with sine function is designed to avoid the boundary problem for rotational box prediction. The advantage of this loss is that the value of loss is also periodic which corresponds exactly to the periodicity of rotational box. Experiment results verify that the detection speed of the proposed method is fastest in comparison with the state-of-the-art methods, while maintaining the competitive detection accuracy. Code is available at https://github.com/ljx43031/Fast-AOOD.
Keyword:
Arbitrary-oriented object detection
boundary problem
YOLOX
期刊
IF:
3.7
论文数:
917
被引数:
2.2K
机构
引用论文
The impact of temperature dependency of the building insulation thermal conductivity in the Canadian climate加拿大气候对建筑保温导热系数温度依赖性的影响
OAAI
OAAI分享收藏Comparative study of layer by layer assembled multilayer films based on graphene oxide and reduced graphene oxide on flexible polyurethane foam: flame retardant and smoke suppression properties.png)
RSC Advances
RSC AdvancesIF0
PREAI
分享收藏Psychometrics of the Child PTSD Symptom Scale for DSM-5 for Trauma-Exposed Children and AdolescentsOAAI
OAAI分享收藏分享收藏
查看更多内容