Article Per Year (5 Year)
p-Index From 2021 - 2026
0.444
P-Index
This Author published in this journals
All Journal IAES International Journal of Artificial Intelligence (IJ-AI) ELINVO (Electronics, Informatics, and Vocational Education)
Danang Wijaya
National Central University
Computer Science & IT Education Electrical & Electronics Engineering Engineering

Published : 2 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 2 Documents
Search

 1 
Optimizing YOLO Models for Enhanced Road Damage Detection: A Performance Comparison of YOLOv5 and YOLOv8 Yuniar Indrihapsari; Danang Wijaya; Satya Adhiyaksa Ardy; Ikhwan Inzaghi Siswanto; Dhista Dwi Nur Ardiansyah; Widya Ardianto
Elinvo (Electronics, Informatics, and Vocational Education) Vol. 10 No. 2 (2025): November 2025
Publisher : Department of Electronic and Informatic Engineering Education, Faculty of Engineering, UNY

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.21831/elinvo.v10i2.88919

Abstract

Accurate road damage detection is vital for ensuring road safety and infrastructure maintenance. This study evaluates and compares the performance of four YOLO models—YOLOv5-S, YOLOv5-M, YOLOv8-S, and YOLOv8-M—for detecting road damage types such as Alligator Cracks, Longitudinal Cracks, Transverse Cracks, Potholes, and Lateral Cracks. The models were trained on a combined dataset from GRDDC 2020 and the Ministry of Public Works and Housing (PUPR) Republic of Indonesia, addressing challenges like class imbalance and diverse road conditions. Results show that YOLOv8-M achieved the highest mAP@0.5 (0.412), excelling in precision and recall for prominent damage types, making it the most reliable for high-accuracy applications. YOLOv5-M balanced precision and recall, while YOLOv5-S prioritized recall, making it suitable for detecting widespread damage. However, all models struggled with less prominent types, such as Lateral Cracks, due to class imbalance. Misclassifications were common, with the "Background" class absorbing predictions from other categories. This study highlights the strengths and limitations of each model, offering insights into improving road damage detection through better feature extraction, expanded datasets, and optimized architectures. These findings provide a foundation for deploying automated deep learning-based road damage detection systems to enhance infrastructure management.
IndoBERT for educational assessment: comparative analysis of transformer models in Indonesian question generation Handaru Jati; Yuniar Indrihapsari; Pradana Setialana; Danang Wijaya; Satya Adhiyaksa Ardy; Dhista Dwi Nur Ardiansyah
IAES International Journal of Artificial Intelligence (IJ-AI) Vol 15, No 2: April 2026
Publisher : Institute of Advanced Engineering and Science

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.11591/ijai.v15.i2.pp1804-1813

Abstract

This study asks whether a monolingual encoder can realistically outperform multilingual and larger transformer models for Indonesian automatic question generation (AQG) when all models share the same training budget. We compare Indonesian bidirectional encoder representations from transformers (IndoBERT), multilingual BERT (mBERT), and BERT-large using a single fine-tuning pipeline with answer highlighting, applied to an Indonesian version of TyDiQA-GoldP and a 20,000 translated subset of SQuAD 2.0. We evaluate model quality using bilingual evaluation understudy score n-gram 4 (BLEU-4), metric for evaluation of translation with explicit ordering (METEOR), and ROUGE-Lincoln (ROUGE-L). IndoBERT consistently achieves the best scores on both datasets (e.g., BLEU-4 of 19.69 on TyDiQA-GoldP and 3.79 on the SQuAD 2.0 subset) while requiring less computation than mBERT and BERT-large. Our results show that language-specific pretraining gives clear advantages for Indonesian AQG, yielding higher accuracy at lower computational cost than multilingual or larger encoders. The work closes a gap in Indonesian AQG benchmarking by providing the first head-to-head comparison of IndoBERT, mBERT, and BERT-large under a shared fine-tuning and evaluation protocol. For educational assessment, the findings offer a practical recipe for building deployable AQG systems on mid-range GPUs that generate higher quality questions without prohibitive training or inference budgets.
Co-Authors Dhista Dwi Nur Ardiansyah Handaru Jati Ikhwan Inzaghi Siswanto Satya Adhiyaksa Ardy Setialana, Pradana Widya Ardianto Yuniar Indrihapsari