Deep Learning for Human Pose Classification using Multi View Dataset
B. Gnana Priya1, M. Arulselvi2
1B. Gnana Priya, Assistant Professor, Department of Computer Science and Engineering, Annamalai University, Chidambaram (Tamil Nadu), India.
2Dr. M. Arulselvi, Assistant Professor, Department of Computer Science and Engineering, Annamalai University, Chidambaram (Tamil Nadu), India.
Manuscript received on 03 June 2019 | Revised Manuscript received on 28 June 2019 | Manuscript Published on 04 July 2019 | PP: 325-328 | Volume-8 Issue-1S4 June 2019 | Retrieval Number: A10570681S419/2019©BEIESP
Open Access | Editorial and Publishing Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Human pose classification isvery challenging area of work in research in modern times. It widely supports in understanding a human poses and its further sequence of actions. Many standard human pose datasets were created and a wide research is taking place. Our main target is to create a multiview dataset containing novel actions which are different from normal poses. Actions from Karate martial arts and Bharathanatyam dance poses are captured. We use Deep Convolutional neural networks to classify the poses without any feature extraction.
Keywords: Action Classification, Deep Learning, Convolutional Neural Network, Karate and Bharathanatyam Dataset.
Scope of the Article: Deep Learning