AI Paper Atlas

Research map / Benchmarks / THuman2.0

THuman2.0 leaderboard

19 papers at ICML, NeurIPS, ICLR, CVPR, AAAI, ICCV and ECCV, and the journals TPAMI, IJCV and JMLR report results on THuman2.0. Best P2S: 0.359 (TPF3D-SIFU-3v, CVPR 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.

THuman2.0 · P2S

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

THuman2.0 · PSNR

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

#PSNR ↑MethodPaperTopic
134.74PFGS (Ours)PFGS: High Fidelity Point Cloud Rendering via Feature Splatting
ECCV 2024 · Table 1
Point cloud classification
228.94RoGSplatRoGSplat: Learning Robust Generalizable Human Gaussian Splatting from Sparse Multi-View Images
CVPR 2025 · Table 1
Gaussian splatting
328.88GM-NeRFGM-NeRF: Learning Generalizable Model-Based Neural Radiance Fields From Multi-View Images
CVPR 2023 · Table: Quantitative results of
NeRF and radiance fields
425.6963HumanRefHumanRef: Single Image to 3D Human Generation via Reference-Guided Diffusion
CVPR 2024 · Table 2
3D generation
525.57GPS-GaussianGPS-Gaussian: Generalizable Pixel-wise 3D Gaussian Splatting for Real-time Human Novel View Synthesis
CVPR 2024 · Table 1
Gaussian splatting
625.1Our Method (Ft)DoubleField: Bridging the Neural Surface and Radiance Fields for High-Fidelity Human Reconstruction and Rendering
CVPR 2022 · Table 2
NeRF and radiance fields
724.033HumanSplatHumanSplat: Generalizable Single-Image Human Gaussian Splatting with Structure Priors
NeurIPS 2024 · Quantitative comparison table
Gaussian splatting
823.54HGM (Ours)Generalizable Human Gaussians from Single-View Image
ICLR 2025 · Novel view synthesis compariso
Face and shape
923.21TPF3D-GTA-3v
GTA
Link to the Past: Temporal Propagation for Fast 3D Human Reconstruction from Monocular Video
CVPR 2025 · Table 1
Face and shape
1021.4ConTex-HumanConTex-Human: Free-View Rendering of Human from a Single Image with Texture-Consistent Synthesis
CVPR 2024 · Table 1
Face and shape

THuman2.0 · SSIM

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

THuman2.0 · LPIPS

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

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