Researchers from University of Science and Technology Report Recent Findings in CDC and FDA (Proposed Methodology for Gait Recognition Using Generative Adversarial Network With Different Feature Selectors).

Researchers from the University of Science and Technology in Egypt have proposed a deep structure procedure for gait recognition using a generative adversarial network (GAN) with different feature selectors. Gait recognition is being investigated as a biometric technology, particularly in light of the COVID-19 pandemic. The researchers used a GAN to generate gait images from pedestrian datasets and extracted gait features using pretrained convolutional neural network (CNN) models. They then balanced the datasets and selected the best features using various feature selectors. The proposed deep neural network achieved high accuracy in identifying different datasets. The research has been peer-reviewed and published in Neural Computing and Applications. [Extracted from the article]