Skip to content
SDGs Center UDINUS

Evidence record · UDINUS · SDGs

Adaptive Inertia Weight Particle Swarm Optimization for Augmentation Selection in Coral Reef Classification with Convolutional Neural Networks

Rektorat UDINUS 2025 Published 2026-10-30

Evidence record · UDINUS · SDGs

Unit
Rektorat UDINUS
Period
2025
Goals
Record
add57d7d-99b9-417a-9d67-62a40a132303
Published
2026-10-30

Universitas Dian Nuswantoro (UDINUS), through the activity participants Prabowo, Dwi Puji; Rohman, Muhammad Syaifur; Megantara, Rama Aria; Pergiwati, Dewi; Saraswati, Galuh Wilujeng; Pramunendar, Ricardus Anggi; Shidik, Guruh Fajar; and Andono, Pulung Nurtantio, conducted a study on augmentation selection for classifying underwater coral reef images using Convolutional Neural Networks (CNNs). The study addresses coral reef image classification in the context of Indonesia, which is located in the coral triangle and has diverse marine life, including vital coral reefs. The source states that coral reefs face threats from climate change, pollution, and human activities, endangering biodiversity and coastal communities. Monitoring and preservation are therefore described as crucial.

The study evaluates various image augmentation methods because underwater coral reef images have unique characteristics. Its methodology tests different augmentation methods, epoch parameters, and CNN parameters on a coral reef image dataset. Five optimization algorithms are compared: Adaptive Inertia Weight Particle Swarm Optimization (AIWPSO), Genetic Algorithm (GA), Grey Wolf Optimizer (GWO), Particle Swarm Optimization (PSO), and FOX. The highest accuracy reported is 95.64%, achieved at the 10th epoch. AIWPSO and GA show the highest average accuracies among the compared algorithms, with 93.44% and 93.50%, respectively. The source reports no significant performance differences among the algorithms.

Statistical analysis using the Wilcoxon test indicates a significant difference between training and validation accuracy, with a p-value of 0.0020. The findings underscore the importance of selecting augmentation methods that align with the characteristics of each optimization algorithm to enhance classification performance. The results provide insights into improving the quality and diversity of input data for classification algorithms in underwater image analysis. They also highlight the necessity of matching augmentation methods to specific optimization algorithms to improve accuracy and effectiveness. The source recommends further exploration of additional augmentation methods and optimization algorithms to enhance the robustness and accuracy of underwater coral reef image classification. The activity is related to SDG 13 and SDG 14 through its source-supported context of climate change threats and underwater coral reef research.

Reference https://www.scopus.com/pages/publications/86000337209?origin=resultslist