metadata

task_categories:
  - image-classification

Dataset for project: food-category-classification

Dataset Description

This dataset is for project food-category-classification.

Languages

The BCP-47 code for the dataset's language is unk.

Dataset Structure

Data Instances

A sample from this dataset looks as follows:

[
  {
    "image": "<512x512 RGB PIL image>",
    "target": 0
  },
  {
    "image": "<512x512 RGB PIL image>",
    "target": 0
  }
]

Dataset Fields

The dataset has the following fields (also called "features"):

{
  "image": "Image(decode=True, id=None)",
  "target": "ClassLabel(names=['Bread', 'Dairy product', 'Dessert', 'Egg', 'Fried food', 'Meat', 'Noodles-Pasta', 'Rice', 'Seafood', 'Soup', 'Vegetable-Fruit'], id=None)"
}

Dataset Splits

This dataset is split into a train and validation split. The split sizes are as follow:

Split name	Num samples
train	1210
valid	275