579 MB
5 files
Updated 6 months ago
Name
Size
.gitattributes778 Bytes
xet
README.md2.39 kB
xet
fairface.py2.07 kB
xet
train.pt514 MB
xet
val.pt65 MB
xet
README.md

Dataset Card for FairFace

Usage

from io import BytesIO
from PIL import Image

import datasets

def bytes_to_pil(example_batch):
    example_batch['img'] = [
        Image.open(BytesIO(b)) for b in example_batch.pop('img_bytes')
    ]
    return example_batch

ds = datasets.load_dataset('nateraw/fairface')
ds = ds.with_transform(bytes_to_pil)

Table of Contents

Dataset Description

Dataset Summary

Existing public face datasets are strongly biased toward Caucasian faces, and other races (e.g., Latino) are significantly underrepresented. This can lead to inconsistent model accuracy, limit the applicability of face analytic systems to non-White race groups, and adversely affect research findings based on such skewed data. To mitigate the race bias in these datasets, we construct a novel face image dataset, containing 108,501 images, with an emphasis of balanced race composition in the dataset. We define 7 race groups: White, Black, Indian, East Asian, Southeast Asian, Middle East, and Latino. Images were collected from the YFCC-100M Flickr dataset and labeled with race, gender, and age groups. Evaluations were performed on existing face attribute datasets as well as novel image datasets to measure generalization performance. We find that the model trained from our dataset is substantially more accurate on novel datasets and the accuracy is consistent between race and gender groups.

Dataset Structure

Data Fields

  • img_bytes: Bytes representing an image
  • age: Age of the person in the image
  • gender: Gender of the person in the image
  • race: Race of the person in the image

Data Instances

{
  'age': 6,
  'gender': 1,
  'img_bytes': b'\\\\\\\\\\\\\\\\xff\\\\\\\\\\\\\\\\xd8...',
  'race': 1
}
Total size
579 MB
Files
5
Last updated
May 28
Pre-warmed CDN
US EU US EU

Contributors