Article Per Year (5 Year)
p-Index From 2021 - 2026
0.817
P-Index
This Author published in this journals
All Journal International Journal of Informatics and Communication Technology (IJ-ICT) Bulletin of Electrical Engineering and Informatics Jurnal Teknimedia: Teknologi Informasi dan Multimedia Journal of Applied Data Sciences
Gede Putra Kusuma
Bina Nusantara University
Computer Science & IT Control & Systems Engineering Decision Sciences, Operations Research & Management Electrical & Electronics Engineering Engineering

Published : 4 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 4 Documents
Search

 1 
Stacking ensemble learning for optical music recognition Francisco Calvin Arnel Ferano; Amalia Zahra; Gede Putra Kusuma
Bulletin of Electrical Engineering and Informatics Vol 12, No 5: October 2023
Publisher : Institute of Advanced Engineering and Science

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.11591/eei.v12i5.5129

Abstract

The development of music culture has resulted in a problem called optical music recognition (OMR). OMR is a task in computer vision that explores the algorithms and models to recognize musical notation. This study proposed the stacking ensemble learning model to complete the OMR task using the common western musical notation (CWMN) musical notation. The ensemble learning model used four deep convolutional neural networks (DCNNs) models, namely ResNeXt50, Inception-V3, RegNetY-400MF, and EfficientNet-V2-S as the base classifier. This study also analysed the most appropriate technique to be used as the ensemble learning model’s meta-classifier. Therefore, several machine learning techniques are determined to be evaluated, namely support vector machine (SVM), logistic regression (LR), random forest (RF), K-nearest neighbor (KNN), decision tree (DT), and Naïve Bayes (NB). Six publicly available OMR datasets are combined, down sampled, and used to test the proposed model. The dataset consists of the HOMUS_V2, Rebelo1, Rebelo2, Fornes, OpenOMR, and PrintedMusicSymbols datasets. The proposed ensemble learning model managed to outperform the model built in the previous study and succeeded in achieving outstanding accuracy and F1-scores with the best value of 97.51% and 97.52%, respectively; both of which were achieved by the LR meta-classifier.
Applying Transfer Learning on Various GNN Model Training in Indoor Positioning System Tasks Kevin Wijaya; Hanif Muhammad Sangga Buana; Gede Putra Kusuma
Journal of Applied Data Sciences Vol 7, No 2: May 2026
Publisher : Bright Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47738/jads.v7i2.1150

Abstract

Determining location and orientation has always been a fundamental challenge, driving advances from maps and compasses to modern global navigation satellite systems (GNSS). However, GNSS performs poorly indoors due to signal attenuation and lack of elevation accuracy, necessitating the development of indoor positioning systems (IPS). Various technologies such as Wi-Fi, Bluetooth Low Energy (BLE), and RFID have been deployed, typically relying on received signal strength (RSS) and fingerprinting to improve accuracy. While previous research focused on training a single model for an entire building, this study explores the creation of floor-specific models by applying transfer learning to various GNN models. This is done to address the substantial signal distortion between floors. Using the UTSIndoorLoc dataset, we evaluate Graph Attention Network (GAT), GraphSAGE, and Graph Convolutional Network (GraphConv) for predicting two-dimensional indoor positions based on RSSI fingerprints. We propose 2 transfer learning model training methods, Schema A and Schema B. Schema A trains the base model iteratively through each floor, and Schema B trains the base model on a unified dataset. Schema B with GraphConv achieved the best results with a mean positioning error of 6.2176 meters. Whilst Schema A achieved a best-case mean positioning error of 6.3900 meters. Both outperforming the standard unified model which has a mean positioning error of 8.0808 meters.
DETEKSI KEPATUHAN ALAT PELINDUNG DIRI DI TERMINAL BAHAN BAKAR MINYAK MENGGUNAKAN CONVOLUTIONAL NEURAL NETWORKS (CNN) Seni Patrisillia; Gede Putra Kusuma
TEKNIMEDIA: Teknologi Informasi dan Multimedia Vol. 6 No. 2 (2025): Desember 2025
Publisher : Badan Penelitian dan Pengabdian Masyarakat (BP2M) STMIK Syaikh Zainuddin NW Anjani

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.46764/teknimedia.v6i2.372

Abstract

Currently, personal protective equipment (PPE) non-compliance is recorded manually through voluntary incident reports, making the data obtained less objective and prone to omissions. This situation drives the need for the application of object detection technology to improve the consistency of monitoring and the accuracy of findings related to PPE compliance. This study aims to compare three object detection architectures—Faster R-CNN, SSD, and YOLOv9—to determine the most suitable model for application in fuel oil terminal environments, particularly in red zone areas that have a high risk of atmospheric explosions. The research method includes collecting 5,000 images from incident reports and public data platforms, followed by annotation, pre-processing, and augmentation to improve model robustness. Faster R-CNN was trained with a ResNet50 and ResNet101 backbone, SSD with the same backbone, while YOLOv9 was tested with variants YOLOv9-C and YOLOv9-E. Model performance was analyzed quantitatively using mean average precision (mAP50) and loss–epoch graphs, and qualitatively through prediction visualization under real-world conditions such as small objects, visual obstructions, and low lighting. The results showed that SSD provided the most stable performance with a balanced mAP50 across training, validation, and test data, as well as consistent visual predictions across various field conditions. In conclusion, the SSD architecture is the most suitable choice for implementing PPE compliance monitoring at fuel terminals due to its acceptable accuracy and robustness to varying environmental conditions.
Leveraging distillation token and weaker teacher model to improve DeiT transfer learning capability Christopher Gavra Reswara; Gede Putra Kusuma
International Journal of Informatics and Communication Technology (IJ-ICT) Vol 15, No 1: March 2026
Publisher : Institute of Advanced Engineering and Science

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.11591/ijict.v15i1.pp198-206

Abstract

Recently, distilling knowledge from convolutional neural networks (CNN) has positively impacted the data-efficient image transformer (DeiT) model. Due to the distillation token, this method is capable of boosting DeiT performance and helping DeiT to learn faster. Unfortunately, a distillation procedure with that token has not yet been implemented in the DeiT for transfer learning to the downstream dataset. This study proposes implementing a distillation procedure based on a distillation token for transfer learning. It boosts DeiT performance on downstream datasets. For example, our proposed method improves the DeiT B 16 model performance by 1.75% on the OxfordIIIT-Pets dataset. Furthermore, we present using a weaker model as a teacher of the DeiT. It could reduce the transfer learning process of the teacher model without reducing the DeiT performance too much. For example, DeiT B 16 model performance decreased by only 0.42% on Oxford 102 Flowers with EfficientNet V2S compared to RegNet Y 16GF. In contrast, in several cases, the DeiT B 16 model performance could improve with a weaker teacher model. For example, DeiT B 16 model performance improved by 1.06% on the OxfordIIIT-Pets dataset with EfficientNet V2S compared to RegNet Y 16GF as a teacher model.
Co-Authors Amalia Zahra Christopher Gavra Reswara Francisco Calvin Arnel Ferano Hanif Muhammad Sangga Buana Kevin Wijaya Seni Patrisillia