Recent study on large scale gait OU-ISIR Treadmill dataset
B with clothing variations, [15] defined three effective body
parts and two redundant parts based on horizontal division
(row wise recognition rate) on DFT representation. They
defined the body into five parts out of which exclude two less
effective parts as redundant and considered only three most
effective body parts for individual recognition. This work
exploits our part based approach using dynamic weighted
entropy based gait representation in our parts definition as
input gait features.