Research Article
Gesture Recognition Controls Image Style Transfer Based on Improved YOLOV5s Algorithm
@INPROCEEDINGS{10.1007/978-3-030-98002-3_15, author={Jiangfan Xie and Huilong Jin and Tian Wen and Ruiyan Du}, title={Gesture Recognition Controls Image Style Transfer Based on Improved YOLOV5s Algorithm}, proceedings={Cognitive Radio Oriented Wireless Networks and Wireless Internet. 16th EAI International Conference, CROWNCOM 2021, Virtual Event, December 11, 2021, and 14th EAI International Conference, WiCON 2021, Virtual Event, November 9, 2021, Proceedings}, proceedings_a={CROWNCOM \& WICON}, year={2022}, month={3}, keywords={Gesture recognition YOLOv5 Human computer interaction}, doi={10.1007/978-3-030-98002-3_15} }- Jiangfan Xie
Huilong Jin
Tian Wen
Ruiyan Du
Year: 2022
Gesture Recognition Controls Image Style Transfer Based on Improved YOLOV5s Algorithm
CROWNCOM & WICON
Springer
DOI: 10.1007/978-3-030-98002-3_15
Abstract
With the rapid development of artificial intelligence, human-computer interaction has drawn more researcher’s attention. As one of the most important ways of human-computer interaction, Gesture recognition has been widely used in many fields. In this paper, an improved YOLOv5s gesture recognition algorithm is proposed, and the results of gesture recognition are used to carry out interactive experiments with the computer. Different gesture selects corresponding style, then the image style transfer network finishes the image style switch according to the image style. At the same time, PyQt5 is used to design an interactive interface to realize gesture recognition and image style conversion. Compared with YOLOv5s, the recall rate of gesture recognition by the improved algorithm is 94.77%, and the average accuracy is 96.46%, and the average accuracy of the improved YOLOv5s is 2.86% higher than YOLOv5s network, which is meeting the requirements of real-time and accuracy of image style transfer.
