Voxel51/Food101

Dataset Card for Food-101

This is a FiftyOne dataset with 35000 samples.

Note: This dataset is subset of the full Food101 dataset. The recipe notebook for creating this dataset can be found here

Installation

If you haven't already, install FiftyOne:

pip install -U fiftyone

Usage

import fiftyone as fo
import fiftyone.utils.huggingface as fouh

# Load the dataset
# Note: other available arguments include 'max_samples', etc
dataset = fouh.load_from_hub("Voxel51/Food101")

# Launch the App
session = fo.launch_app(dataset)

Dataset Details

Dataset Description

The Food-101 dataset is a large-scale dataset for food recognition, consisting of 101,000 images across 101 different food categories.

Here are the key details:

• Contains a total of 101,000 images

• Each food class has 1,000 images, with 750 training images and 250 test images per class

• All images were rescaled to have a maximum side length of 512 pixels

• Curated by: Lukas Bossard, Matthieu Guillaumin, Luc Van Gool

• Funded by: Computer Vision Lab, ETH Zurich, Switzerland

• Shared by: Harpreet Sahota, Hacker-in-Residence at Voxel51

• Language(s) (NLP): en

• License: The dataset images come from Foodspotting and are not owned by the creators of the Food-101 dataset (ETH Zurich). Any use beyond scientific fair use must be negotiated with the respective picture owners according to the Foodspotting terms of use

Dataset Sources

• Repository: https://huggingface.co/datasets/ethz/food101
• Website: https://data.vision.ee.ethz.ch/cvl/datasets_extra/food-101/
• Paper: https://data.vision.ee.ethz.ch/cvl/datasets_extra/food-101/static/bossard_eccv14_food-101.pdf

Citation

BibTeX:

@inproceedings{bossard14,
  title = {Food-101 -- Mining Discriminative Components with Random Forests},
  author = {Bossard, Lukas and Guillaumin, Matthieu and Van Gool, Luc},
  booktitle = {European Conference on Computer Vision},
  year = {2014}
}