Car logo recognition using YOLOv8 and microsoft azure custom vision

Muhammad ‘Arif, Mohd Anuwa and Nor Azuana, Ramli and Mohd Zaid Waqiyuddin, Mohd Zulkifli (2023) Car logo recognition using YOLOv8 and microsoft azure custom vision. In: 4th International Conference on Data Analytics for Business and Industry, ICDABI 2023 , 25 - 27 October 2023 , Bahrain. 477 -481.. ISBN 979-835036978-6 (Published)

	Pdf Car_Logo_Recognition_using_YOLOv8_and_Microsoft_Azure_Custom_Vision.pdf Restricted to Repository staff only Download (1MB) | Request a copy
Preview	Pdf Car logo recognition using YOLOv8_ABST.pdf Download (980kB) | Preview

Abstract

This research is conducted with its main objective to develop an accurate and faster model that can identify brands from logos captured through car images used by Universiti Malaysia Pahang Al-Sultan Abdullah (UMPSA) staff. The software used for this case study is You Only Look Once (YOLO) version 8 and Microsoft Azure's Custom Vision. Each software was compared and results from the analysis showed that YOLOv8 is renowned for its speed and efficiency and is capable of real-time object detection, which makes it ideal for applications where speed is critical. However, this approach might occasionally compromise accuracy, especially for smaller objects or objects that are close together. Microsoft Azure Custom Vision, on the other hand, may not be as fast as YOLOv8, but it generally delivers high accuracy, especially if adequately trained with a diverse set of tagged images. To conclude, the choice between YOLOv8 and Microsoft Azure Custom Vision depends on the specific requirements of the project, technical expertise, and resources.

Item Type:	Conference or Workshop Item (Paper)
Additional Information:	Indexed by Scopus
Uncontrolled Keywords:	Car logo recognition; Image processing; Microsoft Azure; Object recognition; You Only Look Once
Subjects:	Q Science > QA Mathematics > QA75 Electronic computers. Computer science
Faculty/Division:	Center for Mathematical Science Institute of Postgraduate Studies
Depositing User:	Dr. Nor Azuana Ramli
Date Deposited:	06 Sep 2024 03:22
Last Modified:	06 Sep 2024 03:22
URI:	http://umpir.ump.edu.my/id/eprint/42512
Download Statistic:	View Download Statistics

Actions (login required)

View Item