Untrimmed Operator Standard Cleaning Action Parsing Based on Deep Learning Method

Abstract

For the process in the clean room, small particles will not only cause environmental pollution, but also lead to the decrease of product yield. Therefore, it is important to clear away the particles from the body before entering the clean room. This paper described an existing approach for automated monitoring cleaning action on real-time camera. The current method of performing action recognition uses 3D convolutional neural network (3DCNN) and real-time object detection which uses You Only Look Once (YOLO) as backbone. To achieve untrimmed standard cleaning action parsing, our research proposes a new approach by combining the two methods with proposed mechanisms. In addition to considering coarse-grained analysis of different actions, this paper also proposed a fine-grained measure of action completion.