TY - JOUR
T1 - 3DSMDA-Net
T2 - An improved 3DCNN with separable structure and multi-dimensional attention for welding status recognition
AU - Liu, Tianyuan
AU - Wang, Jiacheng
AU - Huang, Xiaodi
AU - Lu, Yuqian
AU - Bao, Jinsong
N1 - Includes bibliographical references
Funding Information:
This work was supported by “the Fundamental Research Funds for the Central Universities and Graduate Student Innovation Fund of Donghua University ” under Grant No. CUSF-DH-D-2020053 .
Publisher Copyright:
© 2021 The Society of Manufacturing Engineers
PY - 2022/1
Y1 - 2022/1
N2 - The vision-based welding status recognition (WSR) provides a basis for online welding quality control. Due to the severe arc and fume interference in the welding area and limited computational resources at the welding edge nodes, it becomes a challenge to mine the most discriminative feature contained in welding images by using a lightweight model. In this paper, we propose an improved three-dimensional convolutional neural network (3DCNN) with separable structure and multi-dimensional attention (3DSMDA-Net) for WSR. The proposed 3DSMDA-Net uses 3DCNN to adaptively extract abstract spatio temporal features in a welding process and then leverages such time sequence information to improve the recognition accuracy of WSR. In addition, we decompose the classical 3D convolution into depth wise convolution and point wise convolution to produce a lightweight model. A multi-dimensional attention mechanism is further proposed to compensate for the loss of accuracy caused by the separation operation. The results of experiments reveal that the proposed method reduces the model size to 1/7 of the classical 3DCNN without sacrificing accuracy. The comparison experiment results have indicated that the accuracy of the proposed method is more accurate and noise-resistant than that of the conventional model.
AB - The vision-based welding status recognition (WSR) provides a basis for online welding quality control. Due to the severe arc and fume interference in the welding area and limited computational resources at the welding edge nodes, it becomes a challenge to mine the most discriminative feature contained in welding images by using a lightweight model. In this paper, we propose an improved three-dimensional convolutional neural network (3DCNN) with separable structure and multi-dimensional attention (3DSMDA-Net) for WSR. The proposed 3DSMDA-Net uses 3DCNN to adaptively extract abstract spatio temporal features in a welding process and then leverages such time sequence information to improve the recognition accuracy of WSR. In addition, we decompose the classical 3D convolution into depth wise convolution and point wise convolution to produce a lightweight model. A multi-dimensional attention mechanism is further proposed to compensate for the loss of accuracy caused by the separation operation. The results of experiments reveal that the proposed method reduces the model size to 1/7 of the classical 3DCNN without sacrificing accuracy. The comparison experiment results have indicated that the accuracy of the proposed method is more accurate and noise-resistant than that of the conventional model.
KW - Arc welding
KW - Status recognition
KW - Deep learning
KW - Time sequence images
KW - 3DCNN
KW - Model lightweight
KW - Multi-dimensional attention
UR - https://www.mendeley.com/catalogue/6c5c4089-68c8-3b3a-a4ca-c6e9031200fe/
UR - http://www.scopus.com/inward/record.url?scp=85100608775&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85100608775&partnerID=8YFLogxK
U2 - 10.1016/j.jmsy.2021.01.017
DO - 10.1016/j.jmsy.2021.01.017
M3 - Article
SN - 0278-6125
VL - 62
SP - 811
EP - 822
JO - Journal of Manufacturing Systems
JF - Journal of Manufacturing Systems
ER -