Categories Companies Research

Loading the SOTA2 catalog…

Explore

Categories Companies Research

Contribute

Join SOTA2 Contact

SOTA2

© 2026 SOTA2

Build in Bellevue, USA

WizWand is now SOTA2 Research

Papers Benchmarks Datasets Tasks

CelebV-HQ dataset and benchmarks · SOTA2 Research

Research Datasets

CelebV-HQ

Benchmarks

Task NameDataset NameSOTA MethodSort

Task NameDataset NameSOTA ResultTrendResults

Video Face Restoration

CelebV-HQ (test)

28.7PSNR

16

Talking Head Generation

CelebV-HQ

0FID

15

Talking Head Generation

CelebV-HQ (test)

6.818FID

11

Talking Avatar Generation

CelebV-HQ (clips)

37.54FID

10

Portrait Image Animation

CelebV-HQ identity-disjoint (test)

7.61FID

8

Video self-reconstruction

CelebV-HQ (test)

0.038L1 Error

8

Talking Head Synthesis

CelebV-HQ (test)

3.588IQA

6

Facial Attribute Recognition

CelebV-HQ

93.9Appearance Accuracy

6

Identity-Preserving Talking Head Generation

CelebV-HQ hard (cross-video)

47.7Speaker Similarity

5

Identity-Preserving Talking Head Generation

CelebV-HQ easy (same-video)

0.573Spk Sim

5

Audio Generation

CelebV-HQ

244.83FAD

5

Unconditional Video Generation

CelebV-HQ 256^2

56.1FVD (16 frames)

5

Image-conditioned text-to-speech

CelebV-HQ

7.938MCD-DTW

2

CelebV-HQ (test)

0.003L1 Loss

2

Face Anonymization

CelebV-HQ

0.011ID Preservation

2

Facial Expression Classification

CelebV-HQ (test)

—Primary metric

0

Popular tasks

Talking Head Generation Identity-Preserving Talking Head Generation Audio Generation Talking head synthesis Facial Attribute Recognition

Follow for updates

Join SOTA2 Contact