Pith. sign in

REVIEW 5 cited by

Generalizable Human Gaussians from Single-View Image

Not yet reviewed by Pith; the record is open.

Re-run · record.json Download PDF Read on arXiv ↗

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2406.06050 v5 pith:4MD7L6LS submitted 2024-06-10 cs.CV

Jinnan Chen , Chen Li , Jianfeng Zhang , Lingting Zhu , Buzhen Huang , Hanlin Chen , Gim Hee Lee This is my paper · ORCID
classification cs.CV
keywords humanimagegaussiansapproachmodelsmpl-xgeneralizableimages
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

In this work, we tackle the task of learning 3D human Gaussians from a single image, focusing on recovering detailed appearance and geometry including unobserved regions. We introduce a single-view generalizable Human Gaussian Model (HGM), which employs a novel generate-then-refine pipeline with the guidance from human body prior and diffusion prior. Our approach uses a ControlNet to refine rendered back-view images from coarse predicted human Gaussians, then uses the refined image along with the input image to reconstruct refined human Gaussians. To mitigate the potential generation of unrealistic human poses and shapes, we incorporate human priors from the SMPL-X model as a dual branch, propagating image features from the SMPL-X volume to the image Gaussians using sparse convolution and attention mechanisms. Given that the initial SMPL-X estimation might be inaccurate, we gradually refine it with our HGM model. We validate our approach on several publicly available datasets. Our method surpasses previous methods in both novel view synthesis and surface reconstruction. Our approach also exhibits strong generalization for cross-dataset evaluation and in-the-wild images.

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 5 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. OASIS: Occlusion-aware Single-image Hand Avatar Reconstruction via 3D Gaussian Splatting

    cs.CV 2026-07 conditional novelty 6.0 of 10

    Single-image hand avatars are reconstructed as animatable 3D Gaussians with visibility-conditioned attention and mesh-guided deformation.

  2. MoGA: 3D Generative Avatar Prior for Monocular Gaussian Avatar Reconstruction

    cs.CV 2025-07 conditional novelty 6.0 of 10

    Fitting a learned 3D Gaussian avatar prior to six diffusion-hallucinated views reconstructs an animatable, high-fidelity avatar from a single image.

  3. PARTE: Part-Guided Texturing for 3D Human Reconstruction from a Single Image

    cs.CV 2025-07 conditional novelty 6.0 of 10

    A part-segmentation-guided diffusion model reconstructs 3D human textures from a single image with improved alignment between textures and human body parts.

  4. Large Images are Gaussians: High-Quality Large Image Representation with Levels of 2D Gaussian Splatting

    cs.CV 2025-02 conditional novelty 6.0 of 10

    A two-level 2D Gaussian splatting method with direct covariance optimization fits large images with more Gaussian points and higher PSNR than prior Gaussian-based image representation.

  5. DevilSight: Augmenting Monocular Human Avatar Reconstruction through a Virtual Perspective

    cs.CV 2025-08 reject novelty 5.0 of 10

    A monocular human avatar reconstruction method generates pseudo back-view videos with a fine-tuned diffusion model and uses them as extra training data for a 3D Gaussian avatar.

Pith tools