Title: Design and development of dynamic gesture recognition system based on deep neural network for driver assistive devices

Authors: Arindam Mondal; Sahadev Roy

Addresses: Department of Electronics and Communication Engineering, National Institute of Technology, Arunachal Pradesh, P.O. Jote, Arunachal Pradesh, 791113, India ' Department of Electronics and Communication Engineering, National Institute of Technology, Arunachal Pradesh, P.O. Jote, Arunachal Pradesh, 791113, India

Abstract: The field of human-computer cooperation is a significant heading in the Internet of Things innovation. Human-Computer cooperation through signals is the bearing of nonstop examination of Internet of Things innovation. The Kinect sensor should be additionally enhanced to acknowledge complex motion developments, particularly the issue that the acknowledgment pace of dynamic signals is not high, which upsets the improvement of human-computer association under the Internet of Things innovation. In work presented in this paper, the Kinect-based motion acknowledgment is investigated exhaustively, and a unique motion acknowledgment strategy dependent on hidden Markov model (HMM) and dynamic-Signal evidence hypothesis is proposed. Based on the first HMM, the digression point and signal change at various snapshots of the palm direction are utilised as the qualities of the perplexing movement motion, and the quantity of quantisation codes diminishes the element of the direction digression.

Keywords: DNN; deep neural network; driver assistive system; D-S evidence theory; gesture recognition; HMM; hidden Markov model.

DOI: 10.1504/IJSSE.2023.133018

International Journal of System of Systems Engineering, 2023 Vol.13 No.3, pp.271 - 283

Received: 12 May 2022
Accepted: 12 Oct 2022
Published online: 24 Aug 2023 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article