Workflow recognition with structured two-stream convolutional networks
Workflow recognition with structured two-stream convolutional networks is a scholarly work, published in 2020 in ''Pattern Recognition Letters''. The main subjects of the publication include database, anomaly detection, Workflow technology, convolutional neural network, factory, data mining, artificial intelligence, timestamp, workflow, process, glossary of archaeology, machine learning, pattern recognition, activity recognition, and computer science. The authors 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.