AI Paper Atlas

Research map / Benchmarks / CelebA 64x64

CelebA 64x64 leaderboard

17 papers at ICML, NeurIPS, ICLR, CVPR, AAAI, ICCV and ECCV, and the journals TPAMI, IJCV and JMLR report results on CelebA 64x64. Best FID: 1.34 (Soft Truncation-G++, ICML 2023).

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 64x64 · FID

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

#FID ↓MethodPaperTopic
11.34Soft Truncation-G++
Soft Truncation
Refining Generative Process with Discriminator Guidance in Score-based Diffusion Models
ICML 2023 · Table: Performance on CelebA/F
Fast diffusion sampling
21.65FaPSFaPS: A General and Fast Training Method for Diffusion Models
ICML 2026 · Table 1
Fast diffusion sampling
31.75INDM (VP, FID)Maximum Likelihood Training of Implicit Nonlinear Diffusion Model
NeurIPS 2022 · Table 4
Fast diffusion sampling
42.1Self-consistency, conditioned guidanceLearn to Guide Your Diffusion Model
ICLR 2026 · Table 2
Fast diffusion sampling
52.62LM (r=30)Likelihood Matching for Diffusion Models
ICML 2026 · Table 1
Fast diffusion sampling
64COCO-GAN (N2,M2,S32)COCO-GAN: Generation by Parts via Conditional Coordinating
ICCV 2019 · Table 1
GANs
74.08ID-GAN+beta-VAEHigh-Fidelity Synthesis with Disentangled Representation
ECCV 2020 · Table 3
GANs
85.01TransGAN converged
TransGAN
TransGAN: Two Pure Transformers Can Make One Strong GAN, and That Can Scale Up
NeurIPS 2021 · Table 3
Attention and transformers
96.19Ours (BW geodesic + adaptive schedule)
BNDM host (U-Net)
Adaptive Noise Covariance Scheduling under Riemannian Metrics for Diffusion Models
ECCV 2026 · Table 1a
Fast diffusion sampling
106.421QAGAN (SSIM)Quality Aware Generative Adversarial Networks
NeurIPS 2019 · Table 2
GANs
7 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.