DFace-Celeb-1M: Benchmarking Face-Hallucination with a Million-scale Dataset

Kaihao Zhang 1, Dongxu Li1, Wenhan Luo2, Jingyu Liu3, Jiankang Deng4, Wei Liu3, Stefanos Zafeiriou4
1Australian National University, 2Sun Yat-sen University, 3Tencent, 4Imperial College London
Paper arXiv Github
Interpolation end reference image.

Sample paired images from the synthetic part of our proposed EDFace-Celeb-1M dataset. The left images are low-resolution face images and the right images are high-resolution ones.

Interpolation end reference image.

Real-world low-resolution images from our proposed EDFace-Celeb-Real dataset.

Abstract

Recent deep face hallucination methods show stunning performance in super-resolving severely degraded facial images, even surpassing human ability. However, these algorithms are mainly evaluated on non-public synthetic datasets. It is thus unclear how these algorithms perform on public face hallucination datasets. Meanwhile, most of the existing datasets do not well consider the distribution of races, which makes face hallucination methods trained on these datasets biased toward some specific races. To address the above two problems, in this paper, we build a public Ethnically Diverse Face dataset, EDFace-Celeb-1M, and design a benchmark task for face hallucination. Our dataset includes 1:7 million photos that cover different countries, with relatively balanced race composition. To the best of our knowledge, it is the largest-scale and publicly available face hallucination dataset in the wild. Associated with this dataset, this paper also contributes various evaluation protocols and provides comprehensive analysis to benchmark the existing state-of-the-art methods. The benchmark evaluations demonstrate the performance and limitations of state-of-the-art algorithms. https://github.com/HDCVLab/EDFace-Celeb-1M

Statistics

Interpolation end reference image.

Dataset

It consists of three sub-datasets, i.e., EDFace-Celeb-1M, EDFace-Celeb-150K and EDFace-Celeb-Real datasets. The EDFace-Celeb-1M and EDFace-Celeb-150K datasets provide different settings for Face Hallucination/Super-resolution (FH128, FH512) and Blind Face Restoration (BFR128, BFR512). We share the links of dataset in both baidu drive and google drive.



Face Hallucination/Super-Resolution Results on FH128.

Method Download Description
DICNet Google Grive Deep Iterative Collaboration Network
DICGAN Google Grive Deep Iterative Collaboration GAN
WaveletSR Google Grive Wavelet-based Super-Resolution Network
HiFaceGAN Google Grive HiFaceGAN
EDSR Google Grive Enhanced deep residual networks
RDN Google Grive Residual Dense Network
RCAN Google Grive Residual Channel Attention Network
HAN Google Grives Holistic Attention Network

Face Hallucination/Super-resolution Results on FH128.

Interpolation end reference image.

Agreement

  • The EDFace-Celeb-1M dataset is only available to download for non-commercial research purposes. The copyright remains with the original owners of the image. A complete version of the license can be found here.
  • You agree not to reproduce, duplicate, copy, sell, trade, resell or exploit for any commercial purposes, any portion of the videos and any portion of derived data. You agree not to further copy, publish or distribute any portion of the EDFace-Celeb-1M dataset.
  • The distribution of identities in the EDFace-Celeb-1M datasets may not be representative of the global human population. Please be careful of unintended societal, gender, racial and other biases when training or deploying models trained on this data.

BibTeX

If you find this helpful, please cite our work: 

@article{zhang2022edface,
        title={EDFace-Celeb-1 M: Benchmarking Face Hallucination with a Million-scale Dataset},
        author={Zhang, Kaihao and Li, Dongxu and Luo, Wenhan and Liu, Jingyu and Deng, Jiankang and Liu, Wei and Zafeiriou, Stefanos},
        journal={IEEE Transactions on Pattern Analysis and Machine Intelligence},
        year={2022}
      }