FFHQ

doi:doi:10.57702/di4n5yi5

FFHQ

Large scale datasets [18, 17, 27, 6] boosted text conditional image generation quality. However, in some domains it could be difficult to make such datasets and usually it could be costly. Also, famous face datasets[7, 11, 13, 29] don’t have corresponding text captions, making it difficult to develop text conditional image generation models on these datasets.

BibTex:

@dataset{Tero_Karras_and_Samuli_Laine_and_Timo_Aila_2024,
    abstract = {Large scale datasets [18, 17, 27, 6] boosted text conditional image generation quality. However, in some domains it could be difficult to make such datasets and usually it could be costly. Also, famous face datasets[7, 11, 13, 29] don’t have corresponding text captions, making it difficult to develop text conditional image generation models on these datasets.},
    author = {Tero Karras and Samuli Laine and Timo Aila},
    doi = {10.57702/di4n5yi5},
    institution = {No Organization},
    keyword = {'3D Images', '3D-consistent synthesis', '3D-controllable', 'Animals', 'Artifact Correction', 'Artifacts', 'Artistic Evaluation', 'Cars', 'Computer Vision', 'Dataset', 'Deep Fake', 'Deep Learning', 'FFHQ', 'Face Dataset', 'Face Editing', 'Face Generation', 'Face Images', 'Face Manipulation', 'Face Recognition', 'Face Reenactment', 'Face Restoration', 'Face dataset', 'Faces', 'Facial Expression', 'Facial Images', 'Fine-grained Control', 'Frequency Analysis', 'GAN', 'GANs', 'Generative Adversarial Networks', 'Generative Models', 'High-Quality Faces', 'High-Quality Images', 'High-Resolution', 'High-Resolution Human Faces', 'High-Resolution Image Generation', 'High-Resolution Images', 'High-quality images', 'High-resolution', 'Human Face', 'Image', 'Image Compression', 'Image Dataset', 'Image Generation', 'Image Recognition', 'Image Super-Resolution', 'Image Synthesis', 'Image synthesis', 'JPEG', 'Object Generation', 'One-Shot', 'Space-filling Curves', 'StyleGAN', 'Text', 'Unconditional Image Generation', 'arbitrary scale', 'artwork style transfer', 'cars', 'cats', 'celebrity images', 'diffusion models', 'eyeglasses manipulation', 'face', 'face images', 'face inpainting', 'face manipulation', 'face recognition', 'face style transfer', 'face synthesis', 'face-to-artwork style transfer', 'faces', 'facial images', 'facial recognition', 'ffhq', 'high-quality faces', 'high-quality images', 'high-resolution images', 'image', 'image dataset', 'image editing', 'image generation', 'image restoration', 'image synthesis', 'image upsampling', 'images', 'monocular images', 'paintings', 'portrait synthesis', 'stylization', 'synthesis', 'unconditional image generation'},
    month = {dec},
    publisher = {TIB},
    title = {FFHQ},
    url = {https://service.tib.eu/ldmservice/dataset/ffhq},
    year = {2024}
}