Implementation Of Hybrid Deep Learning Model For Human Action Recognition

Abstract

In computer vision, human activity recognition is an active research area for different contexts, such as human computer interaction, healthcare, military applications, and security surveillance. Activity recognition is performed to recognize the goals and actions of one or more people from a sequence of observations based on the actions and the environmental conditions. Still, there are numbers of challenges and issues, which motivate the development of new activity recognition method to enhance the accuracy under more realistic conditions. To address the issues, the methodologies proposed in this thesis are: Hybrid Features-Enabled Dragon Deep Belief Neural Network (DDBN), Fuzzy Dragon Deep Belief Neural Network (Fuzzy-DDBN), and Error Based Fuzzy Dragon Deep Belief Neural Network for Activity Recognition using hierarchical skeleton features. newline newlineThe hybrid features-enabled dragon deep belief neural network presents a technique for recognizing human activities in videos using Dragon Deep Belief Network (DDBN) and hybrid features, which comprises of features like shape, coverage factor, and Space-Time Interest (STI) points. The DDBN classifier is designed by the effective combination of DBN and Dragonfly Algorithm. In DDBN, the weights in the network are selected optimally using Dragon Fly Algorithm. The newline newline newline newlineweight update is calculated using the Dragonfly algorithm (DA) for each incoming feature improves the performance of the DDBN classifier. Further it improves the accuracy in classification of actions. From the performance evaluation, the proposed DDBN classifier could attain better performance with 96.01% accuracy, 93.00% sensitivity, and 94.00% specificity. newline newlineFurther, to improve the accuracy, the developed model has been extended by changing the features from hybrid to SIFT and STI, since it outperforms in classification of actions in videos. And also, the design of Deep Neural network has been changed by incorporating fuzzy in to the previous approach. Thus, the extended approach Fuzzy- DDBN classifier has be

Description

Keywords

Citation

item.page.endorsement

item.page.review

item.page.supplemented

item.page.referenced