Adolescent identity search algorithm with optimised video-based activity classification using hierarchical auto-associative polynomial convolutional neural network
by Kaavya Kanagaraj; Shiju George; Asha Joseph; Sushanth H. Gowda
International Journal of Ad Hoc and Ubiquitous Computing (IJAHUC), Vol. 45, No. 4, 2024

Abstract: In this manuscript, video-based activity classification using hierarchical auto-associative polynomial convolutional neural network (V-AC-HA-APCNN) optimised with adolescent identity search algorithm is proposed. Initially, the input action data are taken from Weizmann action dataset. The input data is pre-processed with the help of trilateral filter. Then these pre-processed data are given to force-invariant improved feature extraction (FII-FE) approaches for extracting the necessary features of the video data. These extracted features are given to hierarchical auto-associative polynomial convolutional neural network (HA-APCNN) for classifying the human activities such as walk, run, bend, and skip. Adolescent identity search algorithm (AISA) is considered to enhance the HA-APCNN weight parameters. The performance of the proposed V-AC-HA-APCNN approach attains 32.3%, 56.6%, and 65.5% higher accuracy, and 34.4%, 43.2%, and 32.1% higher ROC compared with existing methods. The intention of this paper is to examine the deep learning methods for the classifications of video-based anomalous activity and focused on anomaly classification.

Online publication date: Wed, 27-Mar-2024

The full text of this article is only available to individual subscribers or to users at subscribing institutions.

 
Existing subscribers:
Go to Inderscience Online Journals to access the Full Text of this article.

Pay per view:
If you are not a subscriber and you just want to read the full contents of this article, buy online access here.

Complimentary Subscribers, Editors or Members of the Editorial Board of the International Journal of Ad Hoc and Ubiquitous Computing (IJAHUC):
Login with your Inderscience username and password:

Username:        Password:          Forgotten your password?

Want to subscribe?
A subscription gives you complete access to all articles in the current issue, as well as to all articles in the previous three years (where applicable). See our Orders page to subscribe.

If you still need assistance, please email subs@inderscience.com