.png){kind=icon}
Return
Workflow recognition with structured two-stream convolutional networks
.png){kind=icon}
OA
AI
DOI：10.1016/j.patrec.2018.10.011.png)
.png){kind=icon}
.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Intelligent monitoring plays an important role in the context of Industry 4.0 '', and behavior recognition is one of the research points in computer vision. However, monitoring the workflow of human beings and machines in production process is very difficult in the real-world complex factory environment. In this paper, we propose a novel workflow recognition framework based on the structured two-stream convolutional neural networks (CNNs) to recognize the behavior of both workers and machines. To improve the accuracy of workflow recognition, we use the CNNs to extract the spatial-temporal features and integrate an attention mechanism to detect the valuable behavior. Then, a Video Triple model is introduced to gain extra timestamp information, which can extend the behavior recognition to workflow recognition. Extensive simulation experiments are conducted on THUMOS'14 dataset and a real-world workflow dataset that show the significant performance improvement in video activity recognition. (C) 2018 Elsevier B.V. All rights reserved.
Keywords:
Workflow recognition
Action recognition
Deep learning
Two-stream CNNs
AI Summary
Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.
Journal
IF:
3.3
Papers:
7.9K
Citations:
1.6W
Organization
Cited Papers
.png){kind=icon}
ShareSaveShareSaveShareSaveShareSave no more
no more