datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
celebamask-hq
Dataset Card for celebamask-hq
This is a FiftyOne dataset with 30000 samples.
Installation
If you haven't already, install FiftyOne:
pip install -U fiftyone
Usage
import fiftyone as fo
from fiftyone.utils.huggingface import load_from_hub
# Load the dataset
# Note: other available arguments include 'max_samples', etc
dataset = load_from_hub("harpreetsahota/celebamask-hq")
# Launch the App
session = fo.launch_app(dataset)
Dataset Details… See the full description on the dataset page: https://huggingface.co/datasets/harpreetsahota/celebamask-hq.CelebA-attrs
CelebA-128x128
CelebA with attrs at 128x128 resolution.
Dataset Information
The attributes are binary attributes. The dataset is already split into train/test/validation sets.
Citation
@inproceedings{liu2015faceattributes,
title = {Deep Learning Face Attributes in the Wild},
author = {Liu, Ziwei and Luo, Ping and Wang, Xiaogang and Tang, Xiaoou},
booktitle = {Proceedings of International Conference on Computer Vision (ICCV)},
month = {December},
year… See the full description on the dataset page: https://huggingface.co/datasets/tpremoli/CelebA-attrs.DigiCam-CelebA-26KData is measured at 30 cm, as shown below.
After downloading and installing LenslessPiCam, the simulated PSF can be obtained and compared with the measured one with the following command:
python scripts/sim/digicam_psf.py \
huggingface_repo=bezzam/DigiCam-CelebA-26K \
sim.waveprop=False \
sim.deadspace=True \
digicam.gamma=2.2 \
digicam.ap_center="[58,76]" \
digicam.ap_shape="[19,25]" \
digicam.rotate=0 \
digicam.horizontal_shift=-60 \
digicam.vertical_shift=-80
For a… See the full description on the dataset page: https://huggingface.co/datasets/bezzam/DigiCam-CelebA-26K.deepfake-celeba
MetaFLOS Deepfake Dataset (CelebA Face Deepfake Generated Images)
Flux2-Klein generated deepfake images from CelebA face descriptions, for deepfake detection / comparison research.
Content
19,867 generated images (full CelebA validation split)
Resolution: 256×256
Each corresponds to a CelebA real face (see real_orig field in manifest)
Generation style: snapshot/crop candid-photo look (not studio portrait) — off-center framing, subject possibly touching or cut by… See the full description on the dataset page: https://huggingface.co/datasets/tjw/deepfake-celeba.celeba-spoof-dataset
Biometric Attack Dataset
The similar dataset that includes all ethnicities - Anti Spoofing Real Dataset
We suggest you the dataset similar to CelebA Dataset but with photos of real people, additionally the dataset for face anti spoofing and face recognition includes not only images, but videos of the individuals!
The videos were gathered by capturing faces of genuine individuals presenting spoofs, using facial presentations. Our dataset proposes a novel approach that… See the full description on the dataset page: https://huggingface.co/datasets/UniqueData/celeba-spoof-dataset.CelebA-attrs-identity
CelebA Dataset from Torchvision
CelebA dataset: https://mmlab.ie.cuhk.edu.hk/projects/CelebA.html
Torchvision documentation: https://docs.pytorch.org/vision/main/generated/torchvision.datasets.CelebA.html
Features
Features include identity (id of the celeb in the image), bbox, landmarks and the 40 binary attributes
Attributes are (0,1) as in the torchvision dataset, not (-1,1) as in the original format
Script generation
traindataset =… See the full description on the dataset page: https://huggingface.co/datasets/negedng/CelebA-attrs-identity.DigiCam-CelebA-10K
Dataset for the paper: https://opg.optica.org/abstract.cfm?uri=pcAOP-2023-JTu4A.45
Data is measured with a computer monitor at 30 cm as shown below (except for the in-the-wild mug measurement which is measured at 12 cm).
After cloning and installing LenslessPiCam, ADMM reconstruction can be applied to the dataset with this script (handles dataset downloading from Hugging Face).python scripts/recon/dataset.py -cn recon_celeba_digicam
The simulated PSF can be obtained and compared with the… See the full description on the dataset page: https://huggingface.co/datasets/bezzam/DigiCam-CelebA-10K.metacloak_celeba_vggface2
Dataset Card for MetaCloak
Dataset Summary
This repository provides datasets from the MetaCloak.
For each dataset, *-gen is the subset used for protecting, and *-eval is used as a clean reference to calculate some quality metrics.
from datasets import load_dataset
dataset = load_dataset("yixin/metacloak_celeba_vggface2")
Contact
Contact Us: yixinliucs@gmail.com
celebahq_512_id_clusters
celebahq_512 with SRK identity labels
Summary
This dataset is a derived version of jxie/celeba-hq. It keeps the original image set and adds automatically generated identity-group labels derived from face-embedding clustering.
As explained in our experimental setup, we use CelebA-HQ from Karras et al. (2018), specifically the Hugging Face snapshot at revision 7ecc6a45edfb5483ccf2f7df1035d298ffe7c76b. The referenced CelebA-HQ version provides gender labels but no identity… See the full description on the dataset page: https://huggingface.co/datasets/edgarcancinoe/celebahq_512_id_clusters.CelebA-HQ-zip
CelebA-HQ dataset:
30k images
1024x1024
.zip archives