PUBLISHED PAPERS #02.05
| Eldar Zeynalli. A Deep Learning Approach for Food Image Classification Using Convolutional Neural Networks |
|---|
| Abstract. Food image classification is a challenging task in computer vision due to the high variability in food appearance, texture, and presentation. This study presents a deep learning-based approach for food image classification using Convolutional Neural Networks (CNNs). We leverage the TensorFlow framework [4] to build and train a CNN model on the Food-101 dataset, achieving state-of-the-art performance. Our results demonstrate the effectiveness of CNNs in handling complex food image datasets, with potential applications in dietary monitoring, food recommendation systems, and automated food logging. The model achieved an accuracy of 85.3% on the test set, with precision and recall scores of 0.86 and 0.85, respectively. We also discuss the challenges of intra-class variability and inter-class similarity, proposing future directions for improving model performance. This work contributes to the growing body of research on deep learning for food image classification, providing insights into the design and optimization of CNN architectures for this task. |
| Keywords: food image classification, CNNs, deep learning, computer vision, image recognition |
 Download PDF |
| DOI: https://doi.org/10.30546/MaCoSEP2025.1090 |