Cover Image

Incremental learning of deep neural network for robust vehicle classification

Existing single-lane free flow (SLFF) tolling systems either heavily rely on contact-based treadle sensor to detect the number of vehicle wheels or manual operator to classify vehicles. While the former is susceptible to high maintenance cost due to wear and tear, the latter is prone to human er...

Full description

Saved in:
Bibliographic Details
Main Authors: Ahmad Mimi Nathiratul Athriyah,, Abdul Kadir Muhammad Amir,, Hasan F. M. Zaki,, Zainal Abidin Zulkifli,, Abdul Rahman Hasbullah,
Format: Article
Language:English
Published: Penerbit Universiti Kebangsaan Malaysia 2022
Online Access:http://journalarticle.ukm.my/20587/1/11.pdf
http://journalarticle.ukm.my/20587/
https://www.ukm.my/jkukm/volume-3405-2022/
Tags: Add Tag
No Tags, Be the first to tag this record!
id my-ukm.journal.20587
record_format eprints
spelling my-ukm.journal.205872022-11-28T12:19:46Z http://journalarticle.ukm.my/20587/ Incremental learning of deep neural network for robust vehicle classification Ahmad Mimi Nathiratul Athriyah, Abdul Kadir Muhammad Amir, Hasan F. M. Zaki, Zainal Abidin Zulkifli, Abdul Rahman Hasbullah, Existing single-lane free flow (SLFF) tolling systems either heavily rely on contact-based treadle sensor to detect the number of vehicle wheels or manual operator to classify vehicles. While the former is susceptible to high maintenance cost due to wear and tear, the latter is prone to human error. This paper proposes a vision-based solution to SLFF vehicle classification by adapting a state-of-the-art object detection model as a backbone of the proposed framework and an incremental training scheme to train our VehicleDetNet in a continual manner to cater the challenging problem of continuous growing dataset in real-world environment. It involved four experiment set-ups where the first stage involved CUTe datasets. VehicleDetNet is utilized for the framework of vehicle detection, and it presents an anchorless network which enable the elimination of the bounding boxes of candidates’ anchors. The classification of vehicles is performed by detecting the vehicle’s location and inferring the vehicle’s class. We augment the model with a wheel detector and enumerator to add more robustness, showing improved performance. The proposed method was evaluated on live dataset collected from the Gombak toll plaza at Kuala Lumpur-Karak Expressway. The results show that within two months of observation, the mean accuracy increases from 87.3 % to 99.07 %, which shows the efficacy of our proposed method. Penerbit Universiti Kebangsaan Malaysia 2022 Article PeerReviewed application/pdf en http://journalarticle.ukm.my/20587/1/11.pdf Ahmad Mimi Nathiratul Athriyah, and Abdul Kadir Muhammad Amir, and Hasan F. M. Zaki, and Zainal Abidin Zulkifli, and Abdul Rahman Hasbullah, (2022) Incremental learning of deep neural network for robust vehicle classification. Jurnal Kejuruteraan, 34 (5). pp. 843-850. ISSN 0128-0198 https://www.ukm.my/jkukm/volume-3405-2022/
institution Universiti Kebangsaan Malaysia
building Tun Sri Lanang Library
collection Institutional Repository
continent Asia
country Malaysia
content_provider Universiti Kebangsaan Malaysia
content_source UKM Journal Article Repository
url_provider http://journalarticle.ukm.my/
language English
description Existing single-lane free flow (SLFF) tolling systems either heavily rely on contact-based treadle sensor to detect the number of vehicle wheels or manual operator to classify vehicles. While the former is susceptible to high maintenance cost due to wear and tear, the latter is prone to human error. This paper proposes a vision-based solution to SLFF vehicle classification by adapting a state-of-the-art object detection model as a backbone of the proposed framework and an incremental training scheme to train our VehicleDetNet in a continual manner to cater the challenging problem of continuous growing dataset in real-world environment. It involved four experiment set-ups where the first stage involved CUTe datasets. VehicleDetNet is utilized for the framework of vehicle detection, and it presents an anchorless network which enable the elimination of the bounding boxes of candidates’ anchors. The classification of vehicles is performed by detecting the vehicle’s location and inferring the vehicle’s class. We augment the model with a wheel detector and enumerator to add more robustness, showing improved performance. The proposed method was evaluated on live dataset collected from the Gombak toll plaza at Kuala Lumpur-Karak Expressway. The results show that within two months of observation, the mean accuracy increases from 87.3 % to 99.07 %, which shows the efficacy of our proposed method.
format Article
author Ahmad Mimi Nathiratul Athriyah,
Abdul Kadir Muhammad Amir,
Hasan F. M. Zaki,
Zainal Abidin Zulkifli,
Abdul Rahman Hasbullah,
spellingShingle Ahmad Mimi Nathiratul Athriyah,
Abdul Kadir Muhammad Amir,
Hasan F. M. Zaki,
Zainal Abidin Zulkifli,
Abdul Rahman Hasbullah,
Incremental learning of deep neural network for robust vehicle classification
author_facet Ahmad Mimi Nathiratul Athriyah,
Abdul Kadir Muhammad Amir,
Hasan F. M. Zaki,
Zainal Abidin Zulkifli,
Abdul Rahman Hasbullah,
author_sort Ahmad Mimi Nathiratul Athriyah,
title Incremental learning of deep neural network for robust vehicle classification
title_short Incremental learning of deep neural network for robust vehicle classification
title_full Incremental learning of deep neural network for robust vehicle classification
title_fullStr Incremental learning of deep neural network for robust vehicle classification
title_full_unstemmed Incremental learning of deep neural network for robust vehicle classification
title_sort incremental learning of deep neural network for robust vehicle classification
publisher Penerbit Universiti Kebangsaan Malaysia
publishDate 2022
url http://journalarticle.ukm.my/20587/1/11.pdf
http://journalarticle.ukm.my/20587/
https://www.ukm.my/jkukm/volume-3405-2022/
_version_ 1751537481652633600
score 13.211869

Similar Items