Journal of Computer Science and Technology

   

Accurate Robotic Grasp Detection with Angular Label Smoothing

Min Shi1 (石敏), Member, CCF, Hao Lu1 (路昊), Zhao-Xin Li2,* (李兆歆), Member, CCF, Deng-Ming Zhu2 (朱登明), Member, CCF, and Zhao-Qi Wang2 (王兆其), Member, CCF, IEEE   

  1. 1School of Control and Computer Engineering, North China Electric Power University, Beijing 1002206, China
    2Institute of Computing Technology, Chinese Academy of Sciences, Beijing 100190, China
  • Contact: Zhao-Xin Li E-mail:lizhaoxin@ict.ac.cn
  • About author:Zhao-Xin Li received the B.Sc. degree and M.Sc. degree in computer science from Tianjin Polytechnic University, Tianjin, in 2007 and 2010, respectively, and received the Ph.D. degree in computer application technology from Harbin Institute of Technology, Harbin, in 2016. From September 2018 to March 2019, he worked as a Postdoctoral Fellow in the Department of Computing, The Hong Kong Polytechnic University. He is currently with the Institute of Computing Technology, Chinese Academy of Sciences, China. His research interests include 3D computer vision and 3D data processing.

Grasp detection is a visual recognition problem where the robot makes use of its sensors to detect graspable objects in its environment. Despite the steady progress in robotic grasping, it is still difficult for the existing methods to achieve both real-time and high accuracy grasping detection. In this paper, we propose a real-time robotic grasp detection method, which can accurately predict potential grasp for parallel-plate robotic gripper using RGB images. Our work employs an end-to-end convolutional neural network which consists of a feature descriptor and a grasp detector, and for the first time, we add an attention mechanism to the grasp detection task, which enables the network to focus on grasp regions rather than background. Specifically, we present an angular label smoothing strategy in our grasp detection method to enhance the fault tolerance of the network. We quantitatively and qualitatively evaluate our grasp detection method from different aspects on the public Cornell dataset and Jacquard dataset. Extensive experiments demonstrate that our grasp detection method achieves superior performance to the state-of-the-art methods. In particular, our grasp detection method ranked No.1 on both the Cornell dataset and the Jacquard dataset, giving rise to accuracy of 98.9% and 95.6% respectively at real-time calculation speed.


中文摘要

1、研究背景(context):
随着人工智能技术的进步,机器人抓取系统得到了相对广泛的研究。机器人抓取工作流程通常可以分为三个连续阶段:抓取检测、轨迹规划和执行抓取。抓取检测任务是一个典型的计算机视觉问题,利用RGB或RGB-D相机等传感器来预测场景中可行的抓取。抓取检测最典型的一种方法是通过生成具有方向的矩形抓取框来预测和表示抓取配置。而作为机器人抓取的第一步,抓取检测的准确性和效率至关重要,直接影响了抓取的性能。
2、目的(Objective):
我们的研究目标是通过研究抓取检测模型的设计以及角度信息的优化,来提出一个更加精准的机器人抓取检测方法,实现从拍摄的RGB图像中实时检测可行的抓取框。
3、方法(Method):
我们提出了一种用于平面抓取预测的新型抓取检测网络,可以实现实时准确地从 RGB 图像中预测五维抓取框。我们采用由特征提取器和抓握检测器组成的端到端卷积神经网络,并且首次将注意力机制引入抓握检测任务中。同时,针对抓取角度信息的学习我们提出了一种用于抓取角度分类的角度标签平滑策略。
4、结果(Result & Findings):
在实时性能方面,我们的模型在 Cornell 数据集和 Jacquard 数据集上分别达到了98.9%和95.6%的准确率,实现了优于目前最先进方法的性能。同时,在实际场景的大量实验表明了所提出的方法具有卓越的性能。
5、结论(Conclusions):
实验结果表明,本文的方法可以通过RGB图像实现快速、鲁棒的抓取预测。同时,我们证明引入的注意力机制和角度标签平滑提高了所提出的抓取检测网络的性能。在未来的工作中,我们将在更复杂的数据集上进一步验证和改进我们的方法,例如更复杂的对象和更嘈杂的场景。此外,我们希望将本文方法能够扩展到不同类型的抓手,以满足更多实际场景的需求。

Key words: robotic grasp detection; attention mechanism; angular label smoothing; anchor box; deep learning in robotic grasping;

[1] Jia-Ke Ge, Yan-Feng Chai, Yun-Peng Chai. WATuning: A Workload-Aware Tuning System with Attention-Based Deep Reinforcement Learning [J]. Journal of Computer Science and Technology, 2021, 36(4): 741-761.
[2] Chen-Chen Sun, De-Rong Shen. Mixed Hierarchical Networks for Deep Entity Matching [J]. Journal of Computer Science and Technology, 2021, 36(4): 822-838.
[3] Yang Liu, Ruili He, Xiaoqian Lv, Wei Wang, Xin Sun, Shengping Zhang. Is It Easy to Recognize Baby's Age and Gender? [J]. Journal of Computer Science and Technology, 2021, 36(3): 508-519.
[4] Yi-Ting Wang, Jie Shen, Zhi-Xu Li, Qiang Yang, An Liu, Peng-Peng Zhao, Jia-Jie Xu, Lei Zhao, Xun-Jie Yang. Enriching Context Information for Entity Linking with Web Data [J]. Journal of Computer Science and Technology, 2020, 35(4): 724-738.
[5] Ying Li, Jia-Jie Xu, Peng-Peng Zhao, Jun-Hua Fang, Wei Chen, Lei Zhao. ATLRec: An Attentional Adversarial Transfer Learning Network for Cross-Domain Recommendation [J]. Journal of Computer Science and Technology, 2020, 35(4): 794-808.
[6] Chun-Yang Ruan, Ye Wang, Jiangang Ma, Yanchun Zhang, Xin-Tian Chen. Adversarial Heterogeneous Network Embedding with Metapath Attention Mechanism [J]. Journal of Computer Science and Technology, 2019, 34(6): 1217-1229.
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
No Suggested Reading articles found!

ISSN 1000-9000(Print)

         1860-4749(Online)
CN 11-2296/TP

Home
Editorial Board
Author Guidelines
Subscription
Journal of Computer Science and Technology
Institute of Computing Technology, Chinese Academy of Sciences
P.O. Box 2704, Beijing 100190 P.R. China
Tel.:86-10-62610746
E-mail: jcst@ict.ac.cn
 
  Copyright ©2015 JCST, All Rights Reserved