Article Per Year (5 Year)
p-Index From 2021 - 2026
0.61
P-Index
This Author published in this journals
All Journal Teknika
Joko Siswantoro
Informatics Engineering, Faculty of Engineering, University of Surabaya, Surabaya, East Java, Indonesia
Computer Science & IT Electrical & Electronics Engineering Engineering

Published : 3 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 3 Documents
Search

 1 
Development of a Mobile Application Using Convolutional Neural Networks for Recognizing Indonesian Traditional Snacks Njoto Benarkah; Joko Siswantoro; Muhammad Ikhsan
Teknika Vol. 14 No. 2 (2025): July 2025
Publisher : Center for Research and Community Service, Institut Informatika Indonesia (IKADO) Surabaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.34148/teknika.v14i2.1236

Abstract

Indonesian traditional snacks constitute a vital element of the country’s cultural heritage. However, growing modernization has contributed to a decline in public familiarity, particularly among younger generations. This study presents a mobile-based image classification desigend to automatically recognize Indonesian traditional snacks using convolutional neural networks (CNNs). A dataset of 3,240 images across 16 snack categories was collected using a smartphone camera. Five CNN architectures, which are, AlexNet, EfficientNetV2M, MobileNetV2, ResNet50V2, and VGG19, were evaluated for classification performance. MobileNetV2 achieved the highest accuracy and F1-score, both reaching 100%. The final model was deployed in a mobile application environment, with the backend developed using Flask and integrated into the Android platform. This research work demonstrates the potential of lightweight CNN models in preserving cultural knowledge through accessible mobile technology.
Vision Transformer-Based Dog Breed Classification with a Hybrid Detection-Classification Framework Njoto Benarkah; Joko Siswantoro; Bryan Porayouw
Teknika Vol. 15 No. 2 (2026): July 2026
Publisher : Center for Research and Community Service, Institut Informatika Indonesia (IKADO) Surabaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.34148/teknika.v15i2.1484

Abstract

Dog breed classification remains a challenging task in computer vision due to high inter-class visual similarity, pose variations, changes in illumination, and complex background conditions. Conventional convolutional neural network (CNN) approaches often struggle to capture global contextual dependencies and subtle discriminative features. This study proposes a hybrid deep learning framework that integrates YOLOv8n for object detection with the Vision Transformer (ViT-B/16) for dog breed classification. The dataset comprises 14,181 dog images collected from the Tsinghua Dogs Dataset and supplementary real-world sources, spanning 10 dog breed categories. The proposed framework includes image preprocessing, data augmentation, transfer learning, and Bayesian hyperparameter optimization using Optuna to enhance model generalization. YOLOv8n is employed to localize dog regions, which are subsequently resized and passed to the Vision Transformer for global feature representation learning. The model is evaluated on 2,133 unseen test images. Experimental results demonstrate that the proposed framework achieves an accuracy of 97.98% with macro and weighted F1-score values of 98.76% and 97.98%, respectively. Comparative experiments against standalone ViT-B/16 and EfficientNetV2M architectures futher confirm the effectiveness of the proposed hybrid YOLOv8n–ViT-B/16 framework for dog breed classification.
A Systematic Literature Review of Convolutional Block Attention Module (CBAM) in Image Classification Vincentius Bernando Wijaya; Joko Siswantoro
Teknika Vol. 15 No. 2 (2026): July 2026
Publisher : Center for Research and Community Service, Institut Informatika Indonesia (IKADO) Surabaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.34148/teknika.v15i2.1500

Abstract

Image classification has become one of the most important tasks in computer vision and has been widely applied in medical imaging, agriculture, surveillance, and industrial systems. Recently, the Convolutional Block Attention Module (CBAM) has attracted significant attention for improving feature representation through channel and spatial attention mechanisms. However, studies on CBAM remain scattered across application domains, backbone architectures, deployment strategies, and evaluation approaches, making its implementation and effectiveness difficult to comprehensively understand. Therefore, this study conducted a Systematic Literature Review (SLR) based on the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) framework to analyze CBAM implementation in image classification studies published between 2018 and 2025. The literature search used Google Scholar through the Publish or Perish application. Of 855 identified articles, 47 met the eligibility criteria and were included. The review analyzed research domains, backbone architectures, CBAM effectiveness, deployment positions, evaluation metrics, and cross-analytical relationships. Medical Imaging was the dominant application domain, while ResNet was the most frequently used backbone architecture. Analysis of 156 ablation experiments showed that 82.1% reported performance improvements after CBAM integration, whereas 91.5% of the reviewed studies demonstrated positive outcomes. Cross-analysis further indicated that CBAM was particularly effective for fine-grained visual recognition tasks, most commonly integrated within backbone architectures, and generally more compatible with residual-based than lightweight architectures. Furthermore, computational efficiency remained underreported. These findings provide a comprehensive synthesis of current CBAM research and practical guidance for future image classification studies.
Co-Authors Bryan Porayouw Muhammad Ikhsan Njoto Benarkah Vincentius Bernando Wijaya