A Confidence-Aware Hybrid Vision–Language Framework for Food Recognition and Nutritional Monitoring
Nutrients · 2026
DOI: 10.3390/nu18152449 ↗Kurz gefasst
Fachartikel in der Zeitschrift Nutrients, 2026. Verfasst von Furkan Göz und 6 weiteren Autor:innen. Beteiligte Einrichtung: Department of Computer Engineering, Kocaeli University, 41001 İzmit, Türkiye. Hauptthemen laut MEDLINE: Nutrition Assessment, Deep Learning, Food. Die Arbeit wurde bisher 2-mal zitiert (Crossref) und nennt 39 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Mälardalen University.
Abstract (Originalsprache)
Background/Objectives: The objective evaluation of regional dietary intake remains a core challenge in personalized health management due to complex plate presentations and a lack of culturally specific dataset benchmarks.
Methods
This study introduces a confidence-aware hybrid vision–language framework engineered for traditional Turkish food recognition and structured nutritional assessment.
Results
We curate a balanced dataset containing 14,711 verified images spanning 40 representative Turkish culinary classes to train and evaluate seven deep learning architectures. Among the visual models, EfficientNet V2-L achieved the highest standalone performance with an accuracy of 93.47%, 0.92 macro-precision, 0.92 macro-recall, and a 0.92 F1 score. To overcome visual ambiguity and automate content analysis, a confidence-aware routing strategy escalates uncertain predictions (τ
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Department of Computer Engineering, Kocaeli University, 41001 İzmit, Türkiye.
- Wireless Information and Intelligent Systems (WINS) Research Center, Kocaeli University, 41001 İzmit, Türkiye.
- Sogang University, Seoul 04107, Republic of Korea.
- Research Center for Interneural Computing, China Medical University, Taichung 40402, Taiwan.
- Centre for Research Impact & Outcome, Chitkara University Institute of Engineering and Technology, Chitkara University, Rajpura 140401, Punjab, India.
- Department of Computer Science and Engineering, Mälardalen University (MDU), 72123 Västerås, Sweden.
ORCID-Profile: Furkan Göz, Muhammad Jamil, Adnan Kavak, Sema Bayraktar, Ali Can Doğru, Gautam Srivastava, Hossein Fotouhi
Förderung
- Mälardalen University
Zitieren
APA
Göz, F., Jamil, M., Kavak, A., Bayraktar, S., Doğru, A. C., Srivastava, G., & Fotouhi, H. (2026). A Confidence-Aware Hybrid Vision–Language Framework for Food Recognition and Nutritional Monitoring. Nutrients. https://doi.org/10.3390/nu18152449
Vancouver
Göz F, Jamil M, Kavak A, Bayraktar S, Doğru AC, Srivastava G, et al. A Confidence-Aware Hybrid Vision–Language Framework for Food Recognition and Nutritional Monitoring. Nutrients. 2026. doi:10.3390/nu18152449
BibTeX
@article{Gz2026,
author = {Göz, Furkan and Jamil, Muhammad and Kavak, Adnan and Bayraktar, Sema and Doğru, Ali Can and Srivastava, Gautam and Fotouhi, Hossein},
title = {A Confidence-Aware Hybrid Vision–Language Framework for Food Recognition and Nutritional Monitoring},
journal = {Nutrients},
year = {2026},
doi = {10.3390/nu18152449}
}
EUVIMED