Geometric Data Augmentation with a Two-Stage Fine-Tuning Strategy for EfficientNetB3-Based Fruit Condition Classification
ZERO: Jurnal Sains, Matematika dan Terapan · 03.04.2026 · 10 (1) · 224
DOI: 10.30829/zero.v10i1.28735 ↗Kurz gefasst
Fachartikel in der Zeitschrift ZERO: Jurnal Sains, Matematika dan Terapan, online erschienen am 03.04.2026. Verfasst von Hana Sajida Azhurra und 2 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY-SA 4.0).
Abstract (Originalsprache)
Accurate fruit condition classification is essential for automated food safety assessment, particularly due to health risks associated with chemical contaminants such as formalin. However, reliable generalization in automated inspection systems remains challenging because limited visual variation in image datasets often leads to overfitting in deep learning models. To address this challenge, this study proposes an EfficientNetB3-based framework that integrates geometric data augmentation with a structured two-stage fine-tuning strategy to improve robustness and training stability. The proposed model achieved 99% test accuracy with consistent cross-dataset performance. The framework also demonstrated stable optimization behavior across training stages, indicating improved generalization capability. From a practical perspective, the proposed approach may support scalable food quality monitoring and automated sorting in agricultural supply chains, as well as preliminary food safety screening in large-scale inspection processes.
Zitieren
APA
Azhurra, H. S., Surono, S., & Thobirin, A. (2026). Geometric Data Augmentation with a Two-Stage Fine-Tuning Strategy for EfficientNetB3-Based Fruit Condition Classification. ZERO: Jurnal Sains, Matematika dan Terapan, 10(1), 224. https://doi.org/10.30829/zero.v10i1.28735
Vancouver
Azhurra HS, Surono S, Thobirin A. Geometric Data Augmentation with a Two-Stage Fine-Tuning Strategy for EfficientNetB3-Based Fruit Condition Classification. ZERO: Jurnal Sains, Matematika dan Terapan. 2026;10(1):224. doi:10.30829/zero.v10i1.28735
BibTeX
@article{Azhurra2026,
author = {Azhurra, Hana Sajida and Surono, Sugiyarto and Thobirin, Aris},
title = {Geometric Data Augmentation with a Two-Stage Fine-Tuning Strategy for EfficientNetB3-Based Fruit Condition Classification},
journal = {ZERO: Jurnal Sains, Matematika dan Terapan},
year = {2026},
volume = {10},
number = {1},
pages = {224},
doi = {10.30829/zero.v10i1.28735}
}
EUVIMED