Papers › PIFuHD: Multi-Level Pixel-Aligned Implicit Function for High-Resolution 3D Human Digitization

PIFuHD: Multi-Level Pixel-Aligned Implicit Function for High-Resolution 3D Human Digitization

1 Apr 2020CVPR 2020 6arXiv:2004.00452archive 2025-07-28

Shunsuke Saito, Tomas Simon, Jason Saragih, Hanbyul Joo

Recent advances in image-based 3D human shape estimation have been driven by the significant improvement in representation power afforded by deep neural networks. Although current approaches have demonstrated the potential in real world settings, they still fail to produce reconstructions with the level of detail often present in the input images. We argue that this limitation stems primarily form two conflicting requirements; accurate predictions require large context, but precise predictions require high resolution. Due to memory limitations in current hardware, previous approaches tend to take low resolution images as input to cover large spatial context, and produce less precise (or low resolution) 3D estimates as a result. We address this limitation by formulating a multi-level architecture that is end-to-end trainable. A coarse level observes the whole image at lower resolution and focuses on holistic reasoning. This provides context to an fine level which estimates highly detailed geometry by observing higher-resolution images. We demonstrate that our approach significantly outperforms existing state-of-the-art techniques on single image human shape reconstruction by fully leveraging 1k-resolution input images.

PaperPDFConference PDFCode

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

Code

facebookresearch/pifuhd officialmentioned on GitHubpytorchNOASSERTION report
agnJason/RGB-D-PIFuHD mentioned on GitHubpytorch report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

3D Human Pose Estimation3D Human Reconstruction3D Human Shape Estimation3D Object Reconstruction From A Single Image3D Shape Reconstruction

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
3D Human Reconstruction 4D-DRESS PIFuHD_Outer Chamfer (cm) 2.393 #9 of 22 Archive leaderboard report
3D Human Reconstruction 4D-DRESS PIFuHD_Outer IoU 0.743 #9 of 22 Archive leaderboard report
3D Human Reconstruction 4D-DRESS PIFuHD_Outer Normal Consistency 0.763 #9 of 22 Archive leaderboard report
3D Human Reconstruction 4D-DRESS PIFuHD_Inner Chamfer (cm) 2.426 #10 of 22 Archive leaderboard report
3D Human Reconstruction 4D-DRESS PIFuHD_Inner IoU 0.739 #10 of 22 Archive leaderboard report
3D Human Reconstruction 4D-DRESS PIFuHD_Inner Normal Consistency 0.793 #10 of 22 Archive leaderboard report
3D Human Reconstruction CAPE PIFuHD Chamfer (cm) 3.237 #3 of 4 Archive leaderboard report
3D Human Reconstruction CAPE PIFuHD NC 0.112 #3 of 4 Archive leaderboard report
3D Human Reconstruction CAPE PIFuHD P2S (cm) 3.123 #3 of 4 Archive leaderboard report
3D Human Reconstruction CustomHumans PIFuHD Chamfer Distance P-to-S 2.107 #3 of 8 Archive leaderboard report
3D Human Reconstruction CustomHumans PIFuHD Chamfer Distance S-to-P 2.228 #3 of 8 Archive leaderboard report
3D Human Reconstruction CustomHumans PIFuHD Normal Consistency 0.804 #3 of 8 Archive leaderboard report
3D Human Reconstruction CustomHumans PIFuHD f-Score 39.076 #3 of 8 Archive leaderboard report
3D Object Reconstruction From A Single Image BUFF ML-PIFu (end-to-end) Chamfer (cm) 1.525 #1 of 5 Archive leaderboard report
3D Object Reconstruction From A Single Image BUFF ML-PIFu (end-to-end) Point-to-surface distance (cm) 0.25 #1 of 5 Archive leaderboard report
3D Object Reconstruction From A Single Image BUFF ML-PIFu (end-to-end) Surface normal consistency 0.22 #1 of 5 Archive leaderboard report
3D Object Reconstruction From A Single Image BUFF ML-PIFu (alternate) Chamfer (cm) 1.73 #5 of 5 Archive leaderboard report
3D Object Reconstruction From A Single Image BUFF ML-PIFu (alternate) Point-to-surface distance (cm) 1.63 #5 of 5 Archive leaderboard report
3D Object Reconstruction From A Single Image BUFF ML-PIFu (alternate) Surface normal consistency 0.133 #5 of 5 Archive leaderboard report
3D Object Reconstruction From A Single Image RenderPeople ML-PIFu (end-to-end) Chamfer (cm) 1.525 #4 of 4 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections