当前位置: X-MOL 学术IEEE Trans. Circ. Syst. Video Technol. › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
Decoupled R-CNN: Sensitivity-Specific Detector for Higher Accurate Localization
IEEE Transactions on Circuits and Systems for Video Technology ( IF 8.4 ) Pub Date : 2022-04-13 , DOI: 10.1109/tcsvt.2022.3167114
Dong Wang , Kun Shang , Huaming Wu , Ce Wang

Object detection, as a fundamental problem in computer vision, has been widely used in many industrial applications, such as intelligent manufacturing and intelligent video surveillance. In this work, we find that classification and regression have different sensitivities to the object translation, from the investigation about the availability of highly overlapping proposals. More specifically, the regressor head has intrinsic characteristics of higher sensitivity to translation than the classifier. Based on it, we propose a decoupled sampling strategy for a deep detector, named Decoupled R-CNN, to decouple the proposals sampling for the two tasks, which induces two sensitivity-specific heads. Furthermore, we adopt the cascaded structure for the single regressor head of Decoupled R-CNN, which is an extremely simple but highly effective way of improving the performance of object detection. Extensive empirical analyses using real-world datasets demonstrate the value of the proposed method when compared with the state-of-the-art models. The reproducing code is available at https://github.com/shouwangzhe134/Decoupled-R-CNN .

中文翻译:

解耦 R-CNN:用于更准确定位的灵敏度特定检测器

目标检测作为计算机视觉中的一个基础问题,已广泛应用于智能制造、智能视频监控等诸多工业应用中。在这项工作中,通过对高度重叠提案的可用性的调查,我们发现分类和回归对对象翻译具有不同的敏感性。更具体地说,回归器头具有比分类器对翻译更敏感的内在特征。在此基础上,我们提出了一种名为 Decoupled R-CNN 的深度检测器的解耦采样策略,以解耦两个任务的提议采样,从而产生两个特定于灵敏度的头部。此外,我们对 Decoupled R-CNN 的单个回归器头采用级联结构,这是提高目标检测性能的一种极其简单但非常有效的方法。与最先进的模型相比,使用真实世界数据集的广泛实证分析证明了所提出方法的价值。复制代码可在https://github.com/shouwangzhe134/Decoupled-R-CNN .
更新日期:2022-04-13
down
wechat
bug