Datasets:

Project-AgML
/

FruitNet_quality_classification

dataset_info:
  features:
    - name: image
      dtype: image
    - name: label
      dtype:
        class_label:
          names:
            '0': Bad
            '1': Good
            '2': Mixed
    - name: crop
      dtype:
        class_label:
          names:
            '0': Apple
            '1': Banana
            '2': Guava
            '3': Lemon
            '4': Lime
            '5': Orange
            '6': Pomegranate
  splits:
    - name: train
      num_bytes: 4009003647
      num_examples: 19526
  download_size: 3291544075
  dataset_size: 4009003647
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
license: cc-by-4.0
task_categories:
  - image-classification
size_categories:
  - 10K<n<100K

Fruitnet Quality Classification

A dataset for image classification of Fruitnet Quality Classification. The dataset contains 19,526 images across 3 classes: Bad, Good, Mixed.
Images per class:

Bad: 6,788
Good: 11,664
Mixed: 1,074

This dataset is indexed on https://project-agml.github.io/ as part of the AgML python library.

Citation

@article{meshram2022fruitnet,
  title={FruitNet: Indian fruits image dataset with quality for machine learning applications},
  author={Meshram, Vishal and Patil, Kailas},
  journal={Data in Brief},
  volume={40},
  pages={107686},
  year={2022},
  publisher={Elsevier}
}

PATIL, Kailas; MESHRAM, Vishal (2022), “FruitNet: Indian Fruits Dataset with quality (Good, Bad & Mixed quality)”, Mendeley Data, V3, doi: 10.17632/b6fftwbr2v.3

This dataset was reformatted from its original format to match HuggingFace standards.