Gait Recognition Based on Feedback Weight Capsule Network

Gait is a change in the posture of a person while walking, and has been widely used in human recognition in recent years. For multi-views of different walking conditions with a small number of markers, inorder to capture robust gait features we propose a gait recognition model based on feedback weighted capsule network. We use a pair of gait energy image(GEI) as input to the network, update the input image with pixel level feedback weights matrix, extract the gait characteristics of the input image with a convolutional network, and output the similarity of the image pairs with the improved capsule network. We experimented with our model on the CASIA-B dataset and the OU-ISIR dataset. The experimental results show that our proposed model performs better than other existing models.