Df-Cbn- A Novel Implementation Ofdeep Learning Framework For An Efficeint Human Gait Analysis

Abstract

Human Gait analysis is well defined technique for human identification and tracking at distance based on their walking style. It plays an important role in the video surveillance, medical and defense applications. However the current gait system might have different dissimilarities due to changing angles and uncertainty features extracted during the gait analysis. Further achieving the higher recognition rate remains as real challenge among the researchers due to varying features of human gaits.Considering the above mentioned problems,convolutional neural network (CNN) is developed to extract the spatial and temporal features from non-invasive camera interfaces. A bidirectional long boosted short term memory (B-LSTM) is integrated  with proposed CNN algorithm in order to detect the level of gaits for an efficient    tracking  the humans .The extensive experiments were carried out with the different datasets and various evaluation parameters such as accuracy, specificity and sensitivity were calculated and results of this work prove to be quite promising, which implies that if the count of databases is increased, then the developed system is able to correctly extract features and appropriately identify humans within a stipulated time.