metadata

pretty_name: Stanford Dogs
license: mit
task_category: Classification

Dataset

This dataset is extracted from Stanford Dogs Dataset

Load

import datasets

dataset = datasets.load_dataset("Alanox/stanford-dogs", split="full")

print(dataset)
"""
Dataset({
    features: ['name', 'annotations', 'target', 'image'],
    num_rows: 20580
})
"""

print(dataset.features)
"""
{
    'name':         Value(dtype='string', id=None),
    'annotations':  Array2D(shape=(None, 4), dtype='int32', id=None),
    # ["xmin", "ymin", "xmax", "ymax"]
    'target':       Value(dtype='string', id=None),
    'image':        Image(decode=True, id=None)
}
"""

This dataset was created by the scripts from this github repo