food_vision_199_classes.py

"""
Loading script for the Food Vision 199 classes dataset.

See the template: https://github.com/huggingface/datasets/blob/main/templates/new_dataset_script.py
See the example for Food101: https://huggingface.co/datasets/food101/blob/main/food101.py 
See another example: https://huggingface.co/datasets/davanstrien/encyclopedia_britannica/blob/main/encyclopedia_britannica.py
"""

import datasets
import os
import requests

import pandas as pd

from datasets.tasks import ImageClassification

# Print datasets version
print(f"Datasets version: {datasets.__version__}")

# Set verbosity to 10
datasets.logging.set_verbosity(10)
print(f"Verbosity level: {datasets.logging.get_verbosity()}")

_HOMEPAGE = "https://www.nutrify.app"
_LICENSE = "TODO"
_CITATION = "TODO"
_DESCRIPTION = "Images of 199 food classes from the Nutrify app."

# # Download class_names.txt and read it
# url = "https://huggingface.co/datasets/mrdbourke/food_vision_199_classes/blob/main/class_names.txt"
# r = requests.get(url, allow_redirects=True)
# open("class_names.txt", "wb").write(r.content)
# with open("class_names.txt", "r") as f:
#     _NAMES = f.read().splitlines()

# Create list of class names
_NAMES = ['almond_butter',
 'almonds',
 'apple',
 'apricot',
 'asparagus',
 'avocado',
 'bacon',
 'bacon_and_egg_burger',
 'bagel',
 'baklava',
 'banana',
 'banana_bread',
 'barbecue_sauce',
 'beans',
 'beef',
 'beef_curry',
 'beef_mince',
 'beef_stir_fry',
 'beer',
 'beetroot',
 'biltong',
 'blackberries',
 'blueberries',
 'bok_choy',
 'bread',
 'broccoli',
 'broccolini',
 'brownie',
 'brussel_sprouts',
 'burrito',
 'butter',
 'cabbage',
 'calamari',
 'candy',
 'capsicum',
 'carrot',
 'cashews',
 'cauliflower',
 'celery',
 'cheese',
 'cheeseburger',
 'cherries',
 'chicken_breast',
 'chicken_thighs',
 'chicken_wings',
 'chilli',
 'chimichurri',
 'chocolate',
 'chocolate_cake',
 'coconut',
 'coffee',
 'coleslaw',
 'cookies',
 'coriander',
 'corn',
 'corn_chips',
 'cream',
 'croissant',
 'crumbed_chicken',
 'cucumber',
 'cupcake',
 'daikon_radish',
 'dates',
 'donuts',
 'dragonfruit',
 'eggplant',
 'eggs',
 'enoki_mushroom',
 'fennel',
 'figs',
 'french_toast',
 'fried_rice',
 'fries',
 'fruit_juice',
 'garlic',
 'garlic_bread',
 'ginger',
 'goji_berries',
 'granola',
 'grapefruit',
 'grapes',
 'green_beans',
 'green_onion',
 'guacamole',
 'guava',
 'gyoza',
 'ham',
 'honey',
 'hot_chocolate',
 'ice_coffee',
 'ice_cream',
 'iceberg_lettuce',
 'jerusalem_artichoke',
 'kale',
 'karaage_chicken',
 'kimchi',
 'kiwi_fruit',
 'lamb_chops',
 'leek',
 'lemon',
 'lentils',
 'lettuce',
 'lime',
 'mandarin',
 'mango',
 'maple_syrup',
 'mashed_potato',
 'mayonnaise',
 'milk',
 'miso_soup',
 'mushrooms',
 'nectarines',
 'noodles',
 'nuts',
 'olive_oil',
 'olives',
 'omelette',
 'onion',
 'orange',
 'orange_juice',
 'oysters',
 'pain_au_chocolat',
 'pancakes',
 'papaya',
 'parsley',
 'parsnips',
 'passionfruit',
 'pasta',
 'pawpaw',
 'peach',
 'pear',
 'peas',
 'pickles',
 'pineapple',
 'pizza',
 'plum',
 'pomegranate',
 'popcorn',
 'pork_belly',
 'pork_chop',
 'pork_loins',
 'porridge',
 'potato_bake',
 'potato_chips',
 'potato_scallop',
 'potatoes',
 'prawns',
 'pumpkin',
 'radish',
 'ramen',
 'raspberries',
 'red_onion',
 'red_wine',
 'rhubarb',
 'rice',
 'roast_beef',
 'roast_pork',
 'roast_potatoes',
 'rockmelon',
 'rosemary',
 'salad',
 'salami',
 'salmon',
 'salsa',
 'salt',
 'sandwich',
 'sardines',
 'sausage_roll',
 'sausages',
 'scrambled_eggs',
 'seaweed',
 'shallots',
 'snow_peas',
 'soda',
 'soy_sauce',
 'spaghetti_bolognese',
 'spinach',
 'sports_drink',
 'squash',
 'starfruit',
 'steak',
 'strawberries',
 'sushi',
 'sweet_potato',
 'tacos',
 'tamarillo',
 'taro',
 'tea',
 'toast',
 'tofu',
 'tomato',
 'tomato_chutney',
 'tomato_sauce',
 'turnip',
 'watermelon',
 'white_onion',
 'white_wine',
 'yoghurt',
 'zucchini']

# Create Food199 class
class Food199(datasets.GeneratorBasedBuilder):
    """Food199 Images dataset"""

    def _info(self):
        return datasets.DatasetInfo(
            description=_DESCRIPTION,
            features=datasets.Features(
                {
                    "image": datasets.Image(),
                    "label": datasets.ClassLabel(names=_NAMES)
                }
            ),
            supervised_keys=("image", "label"),
            homepage=_HOMEPAGE,
            citation=_CITATION,
            license=_LICENSE
        )

    def _split_generators(self, dl_manager):
        """
        This function returns the logic to split the dataset into different splits as well as labels.
        """
        annotations_csv = dl_manager.download("annotations_with_links.csv")
        print(annotations_csv)

        return [
            datasets.SplitGenerator(
                name=datasets.Split.TRAIN,
                gen_kwargs={
                    "annotations": annotations_csv,
                    "split": "train"
                }
            ),
        #     datasets.SplitGenerator(
        #         name=datasets.Split.TEST,
        #         gen_kwargs={
        #             "annotations": annotations_csv,
        #             "split": "test"
        #         }
        #     )
        ]

    def _generate_examples(self, annotations, split):
        """
        This function takes in the kwargs from the _split_generators method and can then yield information from them.
        """
        annotations_df = pd.read_csv(annotations, low_memory=False)

        if split == "train":
            annotations = annotations_df[["image", "label"]][annotations_df["split"] == "train"].to_dict(orient="records")
        elif split == "test":
            annotations = annotations_df[["image", "label"]][annotations_df["split"] == "test"].to_dict(orient="records")

        for id_, row in enumerate(annotations):
            # print(row["image"])
            row["image"] = str(row.pop("image"))
            row["label"] = row.pop("label")
            # print(id_, row)
            yield id_, row