Gian Maxmillian Firdaus and Mahmud Dwi Sulistiy and Hashim, Nik Mohd Zarifie (2024) An improved jellyfish image classification using the EfficientNetB3-Architectured DCNN. In: 2024 12th International Conference on Information and Communication Technology (ICoICT), 7 August 2024 through 8 August 2024, Bandung, Indonesia.
![[img]](http://eprints.utem.edu.my/style/images/fileicons/text.png) |
Text
An Improved Jellyfish Image Classification Using the EfficientNetB3-Architectured DCNN.pdf Restricted to Registered users only Download (714kB) |
Abstract
In recent years, deep learning has significantly advanced the field of computer vision with its ability to process data through multiple nonlinear transformation layers, mirroring the neural network of the human brain. Jellyfish play a crucial role in marine ecosystems, and their population dynamics are indicative of environmental changes, making their classification of utmost importance. Traditional manual classification methods are labor-intensive and require expert knowledge, which necessitates the need for an automated system. This study leverages the power of Deep Convolutional Neural Networks (DCNNs) using the EfficientNetB3 architecture to address the challenge of jellyfish image classification. Prior research has demonstrated success in using various CNN architectures for marine life classification and detection, with some optimizing algorithms for improved detection accuracy. This research employs the EfficientNetB3 model, chosen for its balance between complexity and performance efficiency, and applies it to a dataset of 900 jellyfish images across 6 classes from Kaggle. The model's performance is evaluated based on its classification accuracy, demonstrating that the EfficientNetB3 DCNNs outperform standard CNNs in image classification tasks. The findings suggest that with further refinements in architecture and data domain specialization, such models can offer significant insights and contribute to the field of marine organism monitoring.
| Item Type: | Conference or Workshop Item (Paper) |
|---|---|
| Uncontrolled Keywords: | Jellyfish classification, EfficientNetB3, Deep convolutional neural networks |
| Divisions: | Faculty Of Electronics And Computer Technology And Engineering |
| Depositing User: | NUR FARISAH JAFRIN |
| Date Deposited: | 23 Jul 2026 01:01 |
| Last Modified: | 23 Jul 2026 01:01 |
| URI: | http://eprints.utem.edu.my/id/eprint/29897 |
| Statistic Details: | View Download Statistic |
Actions (login required)
 |
View Item |