Description of Invention
Human gait has unique traits which can be used to distinguish an individual from others, such as stride length, arm swing, limb movement, and etc. Gait is hard to conceal or imitate as the physical properties and body movement differ for each individual. Furthermore, gait recognition requires no cooperation from the subject involved. Gait recognition could be challenging as the walking style can be affected by many factors such as clothing, viewing angle, walking speed and carrying condition. This work presents a multi-model gait recognition that integrates DenseNet-201 model, VGG-16 model and Vision Transformer model.