Gesture recognition based on improved VGGNET convolutional neural network
Yang Zhiqi · 2020 IEEE 5th Information Technology and Mechatronics Engineering Conference (ITOEC) · 2020
In this paper, based on the problem that the existing deep learning neural network relies on large-scale training data and powerful computing power, the convolutional neural network VGGNet is improved, and a modified gesture recognition algorithm based on VGGNet is proposed. Experiments show that under the condition of small training data size and limited computing power, the improved convolutional neural network algorithm has higher recognition rate and can meet the requirements of practical gesture recognition applications.