This repository contains a project focused on classifying images of 101 different food classes using deep learning and computer vision techniques. The project leverages transfer learning and fine-tuning with TensorFlow to achieve a high accuracy of 80% on the full test set, exceeding the baseline accuracy of 50.76% reported in the original Food101 paper using only 10% of the training data, and surpasses the 77.4% accuracy benchmark established by the DeepFood paper using 100% of the training data. The implementation is carried out on Google Colab for efficient computation and easy accessibility.
- Data Source: Preprocessed food image datasets derived from the 'Organizing_The_Food101_Dataset' notebook. Additionally, we are using the same dataset from TensorFlow datasets (TFDS) for some models.
- Deep Learning Framework: TensorFlow
- Model Type: Deep Learning, Computer Vision, Transfer Learning Feature Extraction, and Transfer Learning Fine-Tuning with Data Augmentation
- Goal: Classifying images into 101 different food classes, exceeding the baseline accuracy of 50.76% reported in the original Food101 paper using only 10% of the training data, as well as the 77.4% accuracy benchmark established by the DeepFood paper using 100% of the training data
- Accuracy: Achieves 80% accuracy on the full test set
- Environment: Google Colab for development and execution
- Notebooks:
- Organizing_The_Food101_Dataset: Notebook for preparing and organizing the Food-101 dataset from Kaggle.
- Food Image 101 Classes Classifier - Computer Vision: Notebook detailing the data loading and preprocessing, model training, evaluation, prediction, and results.
- Scripts:
- helper_functions.py: Contains utility functions needed for the project, located in the 'scripts' folder.
- Data: Preprocessed datasets available through publicly accessible Google Drive links in the 'Food Image 101 Classes Classifier - Computer Vision' notebook, along with datasets from TensorFlow Datasets (TFDS).
- Clone the repository:
git clone! https://github.com/IsraelAzoulay/food-image-101-classes-classifier-computer-vision.git- Open the provided Google Colab notebooks:
Navigate to the
notebooksfolder and open the desired notebook in Google Colab. - Run the Notebooks: Follow the instructions in the notebooks to download the datasets, preprocess the data, train, evaluate and predict with the models.
- Customize: Feel free to modify the models and code for your experiments.
Contributions are welcome! Please feel free to submit a Pull Request.
This project is licensed under the Creative Commons Attribution-NonCommercial 4.0 International (CC BY-NC 4.0) License.