AI Paper Atlas

Research map / Benchmarks / CelebA-HQ

CelebA-HQ leaderboard

82 papers at ICML, NeurIPS, ICLR, CVPR, AAAI, ICCV and ECCV, and the journals TPAMI, IJCV and JMLR report results on CelebA-HQ. Best FID: 2.75 (90% noisy + 10% clean (σ=0.2), ICLR 2025).

Numbers are as each paper reports them, read by AI from its results tables and not checked by hand. Papers can differ in setting (training data, model size, extra data); the topic column shows what each paper is about. Open the paper before you compare or quote a number.

CelebA-HQ · FID

65 papers, each with its best reported number for its own method; lower is better.

55 more papers on this leaderboard, the baselines reported on it and a CSV download are free with a scime account.
Sign in free to see all

CelebA-HQ · LPIPS

19 papers, each with its best reported number for its own method; lower is better.

9 more papers on this leaderboard, the baselines reported on it and a CSV download are free with a scime account.
Sign in free to see all

CelebA-HQ · SSIM

18 papers, each with its best reported number for its own method; higher is better.

#SSIM ↑MethodPaperTopic
10.96DTP (λc=1/5)
VGG-19
Deep Translation Prior: Test-Time Training for Photorealistic Style Transfer
AAAI 2022 · Table 1
Editing and style transfer
20.947Ours (EdgeConnect)
EdgeConnect
Image Inpainting With External-Internal Learning and Monochromic Bottleneck
CVPR 2021 · Quantitative comparisons table
Image restoration
30.92DBAEDiffusion Bridge AutoEncoders for Unsupervised Representation Learning
ICLR 2025 · Table 2
Autoregressive and latent generation
40.918Ours
EG3D
High-Fidelity 3D GAN Inversion by Pseudo-Multi-View Optimization
CVPR 2023 · Quantitative evaluation table
Face and shape
50.9DeFLOCNetDeFLOCNet: Deep Image Editing via Flexible Low-Level Controls
CVPR 2021 · Table 1
Editing and style transfer
60.896OursLearning a 3D Morphable Face Reflectance Model From Low-Cost Data
CVPR 2023 · Table: Photometric face recons
Face and shape
70.8676Full attention on structural featuresGenerating Diverse Structure for Image Inpainting With Hierarchical VQ-VAE
CVPR 2021 · Table 2
Autoregressive and latent generation
80.814DiffUEDiffUE: Enhancing Utility-Unlearnability Trade-off of Unlearnable Examples via Diffusion Autoencoders
ECCV 2026 · Table 4
Autoregressive and latent generation
90.811RMT-GANRecoverable Facial Identity Protection via Adaptive Makeup Transfer Adversarial Attacks
AAAI 2025 · Table 2
Attacks on text and graphs
100.8Adv-DiffusionAdv-Diffusion: Imperceptible Adversarial Face Identity Attack via Latent Diffusion Model
AAAI 2024 · Image quality comparison table
Attacks on text and graphs
8 more papers on this leaderboard, the baselines reported on it and a CSV download are free with a scime account.
Sign in free to see all

CelebA-HQ · PSNR

14 papers, each with its best reported number for its own method; higher is better.

4 more papers on this leaderboard, the baselines reported on it and a CSV download are free with a scime account.
Sign in free to see all

Not affiliated with any conference, journal or the benchmark's authors. Corrections: contact@latenticlabs.com.