AntID_APP: A real-time identification software for Taiwanese ants based on the YOLO deep learning model
Journal
IET Conference Proceedings
Journal Volume
2025
Journal Issue
15
Start Page
134
End Page
136
ISSN
27324494
ISBN (of the container)
9781837242634
9781837243143
9781837243150
9781837243167
9781837243235
9781837243341
9781837243358
9781837246847
9781837246854
9781837247004
9781837247011
9781837247028
9781837247035
9781837247042
9781837247271
Date Issued
2025
Author(s)
Abstract
AntID_APP, a new web application, leverages deep learning to provide real-time image analysis and identification of 54 common ant genera in Taiwan. This study details the development process, from image collection and manual annotation using the iNaturalist dataset, through data augmentation and YOLOv9 model training, to the final deployment of the automated identification system. By modularizing the AI image analysis pipeline, this approach reduces development costs and significantly improves object detection efficiency. Unlike traditional biological classification methods, AntID_APP accelerates species identification, freeing up researchers' time and potentially fostering the development of similar systems for other species. The user-friendly website, coupled with a lightweight server architecture, promotes citizen science participation and offers a powerful tool for biodiversity research.
Event(s)
2025 International Conference on Applied System Innovation, ICASI 2025, Tokyo, 22 April 2025 - 25 April 2025
Subjects
ant identification
biodiversity
deep learning
taxonomy identification
YOLOv9
Publisher
Institution of Engineering and Technology
Type
conference paper