Article Per Year (5 Year)
p-Index From 2021 - 2026
0.408
P-Index
This Author published in this journals
All Journal Indonesian Journal of Artificial Intelligence and Data Mining
Bedy Purnama
Telkom University, Bandung, Indonesia
Computer Science & IT

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

Found 2 Documents
Search

 1 
Implementation of EfficientNet-B0-Based Convolutional Neural Network Architecture for Classification of Digital Images of Traditional Spices Muhamad Rafi Raihan Akbar; Bedy Purnama
Indonesian Journal of Artificial Intelligence and Data Mining Vol. 9 No. 2 (2026): July 2026
Publisher : Universitas Islam Negeri Sultan Syarif Kasim Riau

Show Abstract | Download Original | Original Source | Check in Google Scholar

Abstract

Traditional Indonesian spice identification has historically depended on human expertise, a process prone to subjective error and limited scalability. This study evaluates the use of an EfficientNet-B0-based Convolutional Neural Network, incorporating transfer learning and fine-tuning, to automatically classify digital images representing 31 categories of traditional Indonesian spices. The Indonesian Spices Dataset, containing 6,510 images from Kaggle, was divided into 80% training, 10% validation, and 10% testing sets. Data augmentation techniques, such as random horizontal flipping and rotation, were implemented to enhance model generalization and reduce overfitting. The model was trained for over 20 epochs using the AdamW optimizer with cosine learning rate scheduling. Results indicate that the proposed model achieved a test accuracy of 97%, with macro average precision, recall, and F1-score also at 97%. The minimal difference between training and validation accuracy demonstrates robust generalization to unseen data. The model is computationally efficient and suitable for deployment on edge devices, supporting applications in agribusiness for automated spice identification, quality control, and education
Yoga Pose Classification Using Body Landmark-Based Pose Estimation with Mediapipe and Machine Learning Approach Audrey Nasywaa Harimaydina; Bedy Purnama
Indonesian Journal of Artificial Intelligence and Data Mining Vol. 9 No. 2 (2026): July 2026
Publisher : Universitas Islam Negeri Sultan Syarif Kasim Riau

Show Abstract | Download Original | Original Source | Check in Google Scholar

Abstract

Practicing yoga independently without professional supervision can lead to incorrect postures, increasing the risk of injury and reducing exercise effectiveness. Although various studies have utilized pose estimation and machine learning techniques for yoga pose classification, most focus on a single feature representation. This study proposes a static image-based yoga pose classification system using MediaPipe to extract 33 body landmarks, which are transformed into geometric features consisting of landmark coordinates, joint angles, and inter-body point distances. The novelty of this study lies in the systematic evaluation of these features, both individually and in combination, within a Random Forest classification framework. The dataset consists of 1,531 images representing five yoga pose classes: Downward Dog, Goddess, Plank, Tree, and Warrior II. The Random Forest model was optimized using hyperparameter tuning and cross-validation. Experimental results show that combining landmark, angle, and distance features achieved the best performance, with an accuracy of 95.02% and an F1-score of 0.9501. The model also demonstrated stable performance during cross-validation, with accuracy ranging from 0.9436 to 0.9608. The results show that combining multiple geometric feature representations improves yoga pose classification performance while maintaining computational efficiency, supporting safer and more effective self-guided yoga practice
Co-Authors Audrey Nasywaa Harimaydina Muhamad Rafi Raihan Akbar