当前位置: X-MOL 学术Ind. Rob. › 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.)
Accurate grasp detection learning using oriented regression loss
Industrial Robot ( IF 1.8 ) Pub Date : 2021-09-09 , DOI: 10.1108/ir-02-2021-0041
Xuan Zhao 1 , Hancheng Yu 1 , Mingkui Feng 1 , Gang Sun 1
Affiliation  

Purpose

Robot automatic grasping has important application value in industrial applications. Recent works have explored on the performance of deep learning for robotic grasp detection. They usually use oriented anchor boxes (OABs) as detection prior and achieve better performance than previous works. However, the parameters of their loss belong to different coordinates, this may affect the regression accuracy. This paper aims to propose an oriented regression loss to solve the problem of inconsistency among the loss parameters.

Design/methodology/approach

In the oriented loss, the center coordinates errors between the ground truth grasp rectangle and the predicted grasp rectangle rotate to the vertical and horizontal of the OAB. And then the direction error is used as an orientation factor, combining with the errors of the rotated center coordinates, width and height of the predicted grasp rectangle.

Findings

The proposed oriented regression loss is evaluated on the YOLO-v3 framework to the grasp detection task. It yields state-of-the-art performance with an accuracy of 98.8% and a speed of 71 frames per second with GTX 1080Ti on Cornell datasets.

Originality/value

This paper proposes an oriented loss to improve the regression accuracy of deep learning for grasp detection. The authors apply the proposed deep grasp network to the visual servo intelligent crane. The experimental result indicates that the approach is accurate and robust enough for real-time grasping applications.



中文翻译:

使用定向回归损失的准确抓取检测学习

目的

机器人自动抓取在工业应用中具有重要的应用价值。最近的工作探索了深度学习在机器人抓取检测中的性能。他们通常使用定向锚框(OAB)作为先验检测,并获得比以前工作更好的性能。但是,它们的损失参数属于不同的坐标,这可能会影响回归精度。本文旨在提出一种定向回归损失来解决损失参数之间的不一致问题。

设计/方法/方法

在定向损失中,ground truth 抓取矩形和预测抓取矩形之间的中心坐标误差旋转到 OAB 的垂直和水平方向。然后将方向误差作为方向因子,结合旋转后的中心坐标、预测抓取矩形的宽度和高度的误差。

发现

提出的定向回归损失在 YOLO-v3 框架上评估为抓握检测任务。它在康奈尔数据集上使用 GTX 1080Ti 以 98.8% 的准确度和每秒 71 帧的速度产生最先进的性能。

原创性/价值

本文提出了一种定向损失来提高深度学习在抓取检测中的回归精度。作者将所提出的深度抓取网络应用于视觉伺服智能起重机。实验结果表明,该方法对于实时抓取应用来说足够准确和稳健。

更新日期:2021-09-09
down
wechat
bug