arxiv: 2604.21575 · v1 · submitted 2026-04-23 · 💻 cs.CV · cs.GR

Recognition: unknown

OmniFit: Multi-modal 3D Body Fitting via Scale-agnostic Dense Landmark Prediction

Zeyu Cai , Yuliang Xiu , Renke Wang , Zhijing Shao , Xiaoben Li , Siyuan Yu , Chao Xu , Yang Liu

show 3 more authors

Baigui Sun Jian Yang Zhenyu Zhang

Authors on Pith no claims yet

Pith reviewed 2026-05-09 22:29 UTC · model grok-4.3

classification 💻 cs.CV cs.GR

keywords multi-modal 3D body fittingdense landmark predictionscale-agnostictransformer decoderSMPL-Xclothed human reconstructionpartial observations

0 comments

The pith

A conditional transformer decoder maps surface points from multi-modal inputs to dense landmarks for scale-agnostic SMPL-X body fitting.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper presents OmniFit as a way to fit an underlying 3D body model to clothed human data from any combination of full scans, partial depth maps, or single images, without requiring knowledge of the metric scale. Its main innovation is a conditional transformer decoder that converts arbitrary surface points into a set of dense body landmarks. These landmarks then drive the optimization of SMPL-X parameters. An image adapter can be added to supply visual information when geometry is incomplete, and a scale predictor adjusts the subject to standard proportions. This setup is intended to work equally well on real-world captures and AI-generated assets that often have distorted scales.

Core claim

OmniFit shows that directly predicting dense body landmarks with a conditional transformer decoder from diverse surface points allows accurate fitting of the SMPL-X model across different input modalities and without scale information. The decoder conditions on the input points and produces landmarks that serve as targets for parameter estimation. Supporting components include an optional image adapter for visual cues and a scale predictor for canonical proportions. On the CAPE and 4D-DRESS benchmarks the method reaches millimeter-level accuracy while substantially outperforming prior single-modal and multi-view approaches in both daily and loose clothing settings.

What carries the argument

Conditional transformer decoder that maps arbitrary surface points from multi-modal inputs to dense body landmarks for subsequent SMPL-X fitting.

If this is right

Body fitting becomes possible from incomplete or image-only observations.
Scale distortions common in synthetic data no longer prevent accurate reconstruction.
Performance exceeds that of traditional multi-view optimization methods.
Millimeter-level accuracy is achieved on standard clothed body benchmarks.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

This suggests transformer-based decoders could replace hand-crafted correspondence search in other 3D registration tasks.
Future work might test whether the same decoder architecture works for non-human parametric models or dynamic sequences.
Integration with generative image models could allow end-to-end creation of fitted 3D assets from text prompts alone.

Load-bearing premise

The conditional transformer decoder must reliably predict accurate dense landmarks from any combination of surface points and images, even when clothing hides body shape and scale is unknown.

What would settle it

Running the landmark predictor on the CAPE benchmark test set and measuring the resulting SMPL-X fit error; if the average error remains above 5 millimeters, the millimeter-level accuracy claim would be falsified.

Figures

Figures reproduced from arXiv: 2604.21575 by Baigui Sun, Chao Xu, Jian Yang, Renke Wang, Siyuan Yu, Xiaoben Li, Yang Liu, Yuliang Xiu, Zeyu Cai, Zhenyu Zhang, Zhijing Shao.

**Figure 1.** Figure 1: OmniFit handles diverse 3D human data sources and input modalities. (A) Raw point clouds from real-world captures. (B) Point clouds jointly conditioned on a front-view RGB image. (C) Scale-distorted AI-generated 3D human assets. (D) Partial point clouds reconstructed from depth maps. For each case, we display the input point cloud alongside the fitted SMPL-X body, demonstrating the robust fitting capabilit… view at source ↗

**Figure 2.** Figure 2: Landmark Distribution on the SMPL-X Template. We place a total of 600 landmarks across three body regions: 120 for the head, 300 for the body, and 180 for the hands. Point Encoder Capture Point Cloud … … Capture Image Image Encoder 0 1 2 … M-1 Landmark Decoder … MLP Optimization Pred Landmarks Fitted SMPL-X 0 Point Embedding Point Feature Image Feature Landmark Embedding … Optional Frozen [PITH_FULL_IMAGE… view at source ↗

**Figure 3.** Figure 3: The Structure of Landmark Predictor. Given an input point cloud, the landmark predictor estimates dense 3D landmarks. The point cloud is first tokenized into point embeddings and then fed into the point encoder to produce point features. The landmark decoder is a Perceiver-style transformer that takes learnable landmark embeddings as queries, cross-attends to the point cloud features, and regresses the fin… view at source ↗

**Figure 4.** Figure 4: Image Adapter (Left) and Scale Predictor (Right). (Left) The plug-and-play image adapter adds a lightweight cross-attention branch in parallel with the existing point cloud crossattention in each landmark decoder layer. Image features are fused alongside point cloud features without modifying the base architecture, allowing the adapter to be enabled or disabled at will. (Right) The scale predictor shares … view at source ↗

**Figure 5.** Figure 5: Qualitative Comparison with Point Cloud Body Fitting Methods. OmniFit produces more accurate SMPL-X fitting compared to all baselines, especially in the hand and head regions. Failure cases of competing methods are highlighted with red boxes, showing common artifacts such as incorrect hand poses, misaligned head orientations, and inaccurate body shapes. In contrast, OmniFit consistently recovers fine-grain… view at source ↗

**Figure 6.** Figure 6: Comparison with Multi-View Body Fitting. OmniFit with the image adapter outperforms multi-view methods DiffProxy [54] and EasyMocap [1] in body fitting accuracy. Details are highlighted with red boxes and magnified, showing that competing methods struggle with fine-grained regions such as hands, while our adapter recovers these details by incorporating point cloud and image. 4.6 Generalizable Model To impr… view at source ↗

**Figure 7.** Figure 7: Rescaling and Body Fitting Results on Generated 3D Humans. For each example, we show (from left to right) the original 3D human, the rescaled 3D human, and the corresponding SMPL-X fitting result. OmniFit also performs well on scale-distorted human assets. predictor, OmniFit is capable of processing diverse 3D humans produced by generative models. As illustrated in [PITH_FULL_IMAGE:figures/full_fig_p012_7.png] view at source ↗

**Figure 8.** Figure 8: Body Fitting Results with Partial Point Cloud Input. We evaluate OmniFit on partial point cloud inputs. In the right-side case, where entire body regions are missing, OmniFit still produces reasonable SMPL-X fitting results, demonstrating its robustness to incomplete inputs [PITH_FULL_IMAGE:figures/full_fig_p014_8.png] view at source ↗

**Figure 9.** Figure 9: Attention Visualization of Landmarks. We visualize the point cross-attention maps of the landmark decoder. The attention maps accurately reflect part-level correspondence between input surface points and their associated landmarks (Head, Body, Hand). 5 Conclusion We present OmniFit, a unified multi-modal framework for SMPL-X body fitting across diverse types of 3D human assets. OmniFit combines a condition… view at source ↗

**Figure 10.** Figure 10: Visual Examples from Our Two Synthetic Datasets. Left: assets built on BEDLAM2 [50], supplemented with hair and shoe assets. Right: assets built on SynBody [61] driven by poses from Motion-X [32] [PITH_FULL_IMAGE:figures/full_fig_p020_10.png] view at source ↗

**Figure 11.** Figure 11: Partial Data Reconstruction. We simulate single-view point clouds by rendering a 3D asset from a random viewpoint, removing occluded faces and vertices, and sampling points on the remaining surface [PITH_FULL_IMAGE:figures/full_fig_p021_11.png] view at source ↗

**Figure 12.** Figure 12: Body Fitting Results on AI-generated Mesh [PITH_FULL_IMAGE:figures/full_fig_p022_12.png] view at source ↗

**Figure 13.** Figure 13: Fitting Results on Depth Capture. We estimate the depth map from a full-body image using Sapiens [28], extract a point cloud, rescale and fit the 3D body with OmniFit. 3D Human Fitted GT 3D Human Fitted GT [PITH_FULL_IMAGE:figures/full_fig_p023_13.png] view at source ↗

**Figure 14.** Figure 14: Ours vs. Ground-truth. Ground-truth bodies in 4D-DRESS are obtained by optimization, which can get stuck in local minima and produce unreasonable results. OmniFit learns the human body distribution during training, leading to better estimates in som ceases [PITH_FULL_IMAGE:figures/full_fig_p023_14.png] view at source ↗

**Figure 15.** Figure 15 [PITH_FULL_IMAGE:figures/full_fig_p024_15.png] view at source ↗

read the original abstract

Fitting an underlying body model to 3D clothed human assets has been extensively studied, yet most approaches focus on either single-modal inputs such as point clouds or multi-view images alone, often requiring a known metric scale. This constraint is frequently impractical, especially for AI-generated assets where scale distortion is common. We propose OmniFit, a method that can seamlessly handle diverse multi-modal inputs, including full scans, partial depth observations, and image captures, while remaining scale-agnostic for both real and synthetic assets. Our key innovation is a simple yet effective conditional transformer decoder that directly maps surface points to dense body landmarks, which are then used for SMPL-X parameter fitting. In addition, an optional plug-and-play image adapter incorporates visual cues to compensate for missing geometric information. We further introduce a dedicated scale predictor that rescales subjects to canonical body proportions. OmniFit substantially outperforms state-of-the-art methods by 57.1 to 80.9 percent across daily and loose clothing scenarios. To the best of our knowledge, it is the first body fitting method to surpass multi-view optimization baselines and the first to achieve millimeter-level accuracy on the CAPE and 4D-DRESS benchmarks.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

OmniFit uses a conditional transformer to predict dense landmarks from multi-modal inputs without scale, then fits SMPL-X, and reports large gains plus mm accuracy on clothed benchmarks.

read the letter

The key takeaway is that this paper makes scale-agnostic multi-modal 3D body fitting practical by predicting dense landmarks directly with a conditional transformer decoder, then optimizing SMPL-X parameters. It adds a dedicated scale predictor and an optional image adapter for cases with incomplete geometry. The approach handles full scans, partial depth, and images in one pipeline, which addresses a real pain point with AI-generated assets that often lack metric scale.

Referee Report

0 major / 0 minor

Summary. The manuscript introduces OmniFit, a scale-agnostic method for fitting SMPL-X body models to multi-modal 3D inputs including full scans, partial depth maps, and images. The central technical contribution is a conditional transformer decoder that maps arbitrary surface points to dense body landmarks, which are then used for SMPL-X parameter optimization; an optional plug-and-play image adapter supplies visual cues when geometry is incomplete, and a dedicated scale predictor normalizes subjects to canonical proportions. Experiments on the CAPE and 4D-DRESS benchmarks report 57.1–80.9 % error reductions relative to prior art, millimeter-level accuracy, and the first reported outperformance of multi-view optimization baselines under both daily and loose clothing conditions.

Significance. If the reported gains and protocols hold, the work is significant because it removes the metric-scale requirement that limits many existing pipelines, especially for AI-generated assets, while unifying single-view, partial, and multi-modal inputs under one landmark-prediction framework. Credit is due for the explicit experimental protocols, the internal consistency of the method (no circular reliance on self-generated data), and the reproducible benchmark leadership claims on CAPE and 4D-DRESS.

Simulated Author's Rebuttal

0 responses · 0 unresolved

We thank the referee for their positive assessment of the manuscript, recognition of its technical contributions in scale-agnostic multi-modal fitting, and recommendation for minor revision. We appreciate the acknowledgment of the experimental protocols, internal consistency, and benchmark results on CAPE and 4D-DRESS.

Circularity Check

0 steps flagged

No significant circularity in derivation chain

full rationale

The paper's core pipeline—conditional transformer decoder predicting dense landmarks from multi-modal surface points, followed by standard SMPL-X parameter optimization and an independent scale predictor module—is presented as a sequence of trained components validated on external benchmarks (CAPE, 4D-DRESS). No step reduces by construction to its own inputs, fitted parameters renamed as predictions, or load-bearing self-citations; the landmark-to-SMPL-X fitting uses conventional optimization rather than self-referential outputs, and scale handling is a dedicated separate module without ansatz smuggling or uniqueness claims imported from prior author work.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 2 invented entities

The central claim depends on the unproven effectiveness of the newly introduced conditional transformer decoder and scale predictor for landmark accuracy across input types. Standard domain assumptions about the SMPL-X model are invoked without new evidence.

axioms (1)

domain assumption The SMPL-X parametric model is a sufficiently accurate representation for fitting to clothed human geometry from multi-modal inputs
All fitting is performed by optimizing SMPL-X parameters from the predicted landmarks.

invented entities (2)

Conditional transformer decoder for dense landmark prediction no independent evidence
purpose: Directly maps surface points from multi-modal inputs to dense body landmarks
Presented as the core innovation enabling scale-agnostic operation.
Dedicated scale predictor no independent evidence
purpose: Rescales input subjects to canonical body proportions
Introduced to address scale ambiguity in the inputs.

pith-pipeline@v0.9.0 · 5548 in / 1473 out tokens · 36126 ms · 2026-05-09T22:29:47.015782+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

69 extracted references · 5 canonical work pages · 1 internal anchor

[1]

Github (2021),https://github

EasyMoCap - Make human motion capture easier. Github (2021),https://github. com/zju3dv/EasyMocap

2021
[2]

Transactions on Graphics (TOG) (2003)

Allen, B., Curless, B., Popović, Z.: The space of human body shapes: reconstruction and parameterization from range scans. Transactions on Graphics (TOG) (2003)

2003
[3]

In: International Conference on 3D Vision (3DV) (2024)

Antić, D., Tiwari, G., Ozcomlekci, B., Marin, R., Pons-Moll, G.: Close: a 3D clothing segmentation dataset and model. In: International Conference on 3D Vision (3DV) (2024)

2024
[4]

In: European Conference on Computer Vision (ECCV) (2020)

Bertiche, H., Madadi, M., Escalera, S.: Cloth3d: Clothed 3d humans. In: European Conference on Computer Vision (ECCV) (2020)

2020
[5]

Bhatnagar, B., Petrov, I., Xie, X.: RVH mesh registration.https://github.com/ bharat-b7/RVH_Mesh_Registration(2022)

2022
[6]

In: European Conference on Computer Vision (ECCV) (2020)

Bhatnagar, B.L., Sminchisescu, C., Theobalt, C., Pons-Moll, G.: Combining implicit function learning and parametric models for 3d human reconstruction. In: European Conference on Computer Vision (ECCV) (2020)

2020
[7]

In: Conference on Neural Information Processing Systems (NeurIPS) (2020)

Bhatnagar, B.L., Sminchisescu, C., Theobalt, C., Pons-Moll, G.: Loopreg: self- supervised learning of implicit surface correspondences, pose and shape for 3d human mesh registration. In: Conference on Neural Information Processing Systems (NeurIPS) (2020)

2020
[8]

In: IEEE Conference on Artificial Intelligence (CAI) (2024)

Cai, Z., Huang, Z., Zheng, X., Liu, Y., Liu, C., Wang, Z., Wang, L.: Interact360: Interactive identity-driven text to 360 panorama generation. In: IEEE Conference on Artificial Intelligence (CAI) (2024)

2024
[9]

In: International Conference on Learning Representations (ICLR) (2026)

Cai, Z., Li, Z., Li, X., Li, B., Wang, Z., Zhang, Z., Xiu, Y.: Up2you: Fast reconstruc- tion of yourself from unconstrained photo collections. In: International Conference on Learning Representations (ICLR) (2026)

2026
[10]

arXiv preprint arXiv:2409.05099 (2024) 5

Cai, Z., Wang, D., Liang, Y., Shao, Z., Chen, Y.C., Zhan, X., Wang, Z.: Dreammap- ping: High-fidelity text-to-3d generation via variational distribution mapping. arXiv preprint arXiv:2409.05099 (2024)

work page arXiv 2024
[11]

In: IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2024)

Cao, Y., Cao, Y.P., Han, K., Shan, Y., Wong, K.Y.K.: Dreamavatar: Text-and-shape guided 3d human avatar generation via diffusion models. In: IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2024)

2024
[12]

Transactions on Pattern Analysis and Machine Intelligence (TPAMI) (2019)

Cao, Z., Hidalgo Martinez, G., Simon, T., Wei, S., Sheikh, Y.A.: Openpose: Realtime multi-person 2d pose estimation using part affinity fields. Transactions on Pattern Analysis and Machine Intelligence (TPAMI) (2019)

2019
[13]

In: IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2025)

Chen, J., Hu, J., Wang, G., Jiang, Z., Zhou, T., Chen, Z., Lv, C.: Taoavatar: Real- time lifelike full-body talking avatars for augmented reality via 3d gaussian splatting. In: IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2025)

2025
[14]

Conference on Neural Information Processing Systems (NeurIPS) (2025)

Chen, W., Li, P., Zheng, W., Zhao, C., Li, M., Zhu, Y., Dou, Z., Wang, R., Liu, Y.: SyncHuman: Synchronizing 2d and 3d generative models for single-view human reconstruction. Conference on Neural Information Processing Systems (NeurIPS) (2025)

2025
[15]

Image and vision computing (1992)

Chen, Y., Medioni, G.: Object modelling by registration of multiple range images. Image and vision computing (1992)

1992
[16]

MAMMA: Markerless & Automatic Multi-Person Motion Action Capture

Cuevas-Velasquez, H., Yiannakidis, A., Shin, S., Becherini, G., Höschle, M., Tesch, J., Obersat, T., Alexiadis, T., Halilaj, E., Black, M.J.: Mamma: Markerless & automatic multi-person motion action capture. arXiv preprint arXiv:2506.13040 (2025)

work page internal anchor Pith review Pith/arXiv arXiv 2025
[17]

Transactions on Pattern Analysis and Machine Intelligence (TPAMI) (2022) 16 OmniFit arXiv Preprint

Fang, H.S., Li, J., Tang, H., Xu, C., Zhu, H., Xiu, Y., Li, Y.L., Lu, C.: Alpha- pose: Whole-body regional multi-person pose estimation and tracking in real-time. Transactions on Pattern Analysis and Machine Intelligence (TPAMI) (2022) 16 OmniFit arXiv Preprint

2022
[18]

In: International Conference on Computer Vision (ICCV) (2023)

Feng, H., Kulits, P., Liu, S., Black, M.J., Abrevaya, V.F.: Generalizing neural human fitting to unseen poses with articulated se (3) equivariance. In: International Conference on Computer Vision (ICCV) (2023)

2023
[19]

In: IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2019)

Gong, K., Gao, Y., Liang, X., Shen, X., Wang, M., Lin, L.: Graphonomy: Universal human parsing via graph transfer learning. In: IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2019)

2019
[20]

In: IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2023)

Han, S.H., Park, M.G., Yoon, J.H., Kang, J.M., Park, Y.J., Jeon, H.G.: High-fidelity 3d human digitization from single 2k resolution images. In: IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2023)

2023
[21]

In: International Conference on Learning Representa- tions (ICLR) (2025)

He, C., Sun, X., Shu, Z., Luan, F., Pirk, S., Herrera, J.A.A., Michels, D.L., Wang, T.Y., Zhang, M., Rushmeier, H., Zhou, Y.: Perm: A parametric representation for multi-style 3D hair modeling. In: International Conference on Learning Representa- tions (ICLR) (2025)

2025
[22]

In: AAAI Conference on Artificial Intelligence (2025)

He, X., Wu, Z., Li, X., Kang, D., Zhang, C., Ye, J., Chen, L., Gao, X., Zhang, H., Zhuang, H.: Magicman: Generative novel view synthesis of humans with 3d-aware diffusion and iterative refinement. In: AAAI Conference on Artificial Intelligence (2025)

2025
[23]

In: IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2023)

Ho, H.I., Xue, L., Song, J., Hilliges, O.: Learning locally editable virtual humans. In: IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2023)

2023
[24]

In: International Conference on Computer Vision (ICCV) (2025)

Huang, Z., Guo, Y.C., Wang, H., Yi, R., Ma, L., Cao, Y.P., Sheng, L.: Mv-adapter: Multi-view consistent image generation made easy. In: International Conference on Computer Vision (ICCV) (2025)

2025
[25]

In: International Conference on Machine Learning (ICML) (2021)

Jaegle, A., Gimeno, F., Brock, A., Vinyals, O., Zisserman, A., Carreira, J.: Perceiver: General perception with iterative attention. In: International Conference on Machine Learning (ICML) (2021)

2021
[26]

In: International Conference on Computer Vision (ICCV) (2019)

Jiang, H., Cai, J., Zheng, J.: Skeleton-aware 3d human shape reconstruction from point clouds. In: International Conference on Computer Vision (ICCV) (2019)

2019
[27]

Transactions on Graphics (TOG) (2023)

Kerbl, B., Kopanas, G., Leimkühler, T., Drettakis, G., et al.: 3d gaussian splatting for real-time radiance field rendering. Transactions on Graphics (TOG) (2023)

2023
[28]

In: European Conference on Computer Vision (ECCV) (2024)

Khirodkar, R., Bagautdinov, T., Martinez, J., Zhaoen, S., James, A., Selednik, P., Anderson, S., Saito, S.: Sapiens: Foundation for human vision models. In: European Conference on Computer Vision (ECCV) (2024)

2024
[29]

arXiv preprint arXiv:2504.03602 (2025)

Lascheit, K., Barath, D., Pollefeys, M., Guibas, L., Engelmann, F.: Robust human registration with body part segmentation on noisy point clouds. arXiv preprint arXiv:2504.03602 (2025)

work page arXiv 2025
[30]

In: International Conference on Computer Vision (ICCV) (2025)

Li, B., Feng, H., Cai, Z., Black, M.J., Xiu, Y.: Etch: Generalizing body fitting to clothed humans via equivariant tightness. In: International Conference on Computer Vision (ICCV) (2025)

2025
[31]

In: IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2025)

Li, P., Zheng, W., Liu, Y., Yu, T., Li, Y., Qi, X., Chi, X., Xia, S., Cao, Y.P., Xue, W., et al.: Pshuman: Photorealistic single-image 3d human reconstruction using cross-scale multiview diffusion and explicit remeshing. In: IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2025)

2025
[32]

Conference on Neural Information Processing Systems (NeurIPS) (2023)

Lin, J., Zeng, A., Lu, S., Cai, Y., Zhang, R., Wang, H., Zhang, L.: Motion-x: A large-scale 3d expressive whole-body human motion dataset. Conference on Neural Information Processing Systems (NeurIPS) (2023)

2023
[33]

In: ACM International Conference on Multimedia (MM) (2021)

Liu, G., Rong, Y., Sheng, L.: Votehmr: Occlusion-aware voting network for robust 3d human mesh recovery from partial point clouds. In: ACM International Conference on Multimedia (MM) (2021)

2021
[34]

In: International Conference on Computer Vision (ICCV) (2020) OmniFit arXiv Preprint 17

Ma, Q., Yang, J., Ranjan, A., Pujades, S., Pons-Moll, G., Tang, S., Black, M.J.: Learning to dress 3d people in generative clothing. In: International Conference on Computer Vision (ICCV) (2020) OmniFit arXiv Preprint 17

2020
[35]

In: International Conference on Computer Vision (ICCV) (2019)

Mahmood, N., Ghorbani, N., Troje, N.F., Pons-Moll, G., Black, M.J.: Amass: Archive of motion capture as surface shapes. In: International Conference on Computer Vision (ICCV) (2019)

2019
[36]

In: European Conference on Computer Vision (ECCV) (2024)

Marin, R., Corona, E., Pons-Moll, G.: Nicp: Neural icp for 3d human registration at scale. In: European Conference on Computer Vision (ECCV) (2024)

2024
[37]

In: IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2021)

Müller, L., Osman, A.A.A., Tang, S., Huang, C.H.P., Black, M.J.: On self-contact and human pose. In: IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2021)

2021
[38]

Transactions on Machine Learning Research Journal (TMLR) (2024)

Oquab, M., Darcet, T., Moutakanni, T., Vo, H., Szafraniec, M., Khalidov, V., Fernandez, P., Haziza, D., Massa, F., El-Nouby, A., et al.: Dinov2: Learning robust visual features without supervision. Transactions on Machine Learning Research Journal (TMLR) (2024)

2024
[39]

In: European Conference on Computer Vision (ECCV) (2024)

Örnek, E.P., Labbé, Y., Tekin, B., Ma, L., Keskin, C., Forster, C., Hodan, T.: Foundpose: Unseen object pose estimation with foundation features. In: European Conference on Computer Vision (ECCV) (2024)

2024
[40]

In: IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2021)

Patel, P., Huang, C.H.P., Tesch, J., Hoffmann, D.T., Tripathi, S., Black, M.J.: Agora: Avatars in geography optimized for regression analysis. In: IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2021)

2021
[41]

In: IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2019)

Pavlakos, G., Choutas, V., Ghorbani, N., Bolkart, T., Osman, A.A.A., Tzionas, D., Black, M.J.: Expressive body capture: 3d hands, face, and body from a single image. In: IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2019)

2019
[42]

Transactions on Graphics (TOG) (2015)

Pons-Moll, G., Romero, J., Mahmood, N., Black, M.J.: Dyna: A model of dynamic human shape in motion. Transactions on Graphics (TOG) (2015)

2015
[43]

In: IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2017)

Qi, C.R., Su, H., Mo, K., Guibas, L.J.: Pointnet: Deep learning on point sets for 3d classification and segmentation. In: IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2017)

2017
[44]

In: Conference on Neural Information Processing Systems (NeurIPS) (2017)

Qi, C.R., Yi, L., Su, H., Guibas, L.J.: Pointnet++: Deep hierarchical feature learning on point sets in a metric space. In: Conference on Neural Information Processing Systems (NeurIPS) (2017)

2017
[45]

In: International Conference on Computer Vision (ICCV) (2025)

Qiu, L., Gu, X., Li, P., Zuo, Q., Shen, W., Zhang, J., Qiu, K., Yuan, W., Chen, G., Dong, Z., et al.: Lhm: Large animatable human reconstruction model from a single image in seconds. In: International Conference on Computer Vision (ICCV) (2025)

2025
[46]

In: IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2025)

Qiu, L., Zhu, S., Zuo, Q., Gu, X., Dong, Y., Zhang, J., Xu, C., Li, Z., Yuan, W., Bo, L., et al.: Anigs: Animatable gaussian avatar from a single image with inconsistent gaussian reconstruction. In: IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2025)

2025
[47]

Transactions on Graphics (TOG) (2024)

Shao, R., Pang, Y., Zheng, Z., Sun, J., Liu, Y.: 360-degree human video generation with 4d diffusion transformer. Transactions on Graphics (TOG) (2024)

2024
[48]

In: International Conference on 3D Vision (3DV) (2025)

Shao, Z., Wang, D., Tian, Q.Y., Yang, Y.D., Meng, H., Cai, Z., Dong, B., Zhang, Y., Zhang, K., Wang, Z.: Degas: Detailed expressions on full-body gaussian avatars. In: International Conference on 3D Vision (3DV) (2025)

2025
[49]

In: IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2023)

Shen, K., Guo, C., Kaufmann, M., Zarate, J.J., Valentin, J., Song, J., Hilliges, O.: X-avatar: Expressive human avatars. In: IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2023)

2023
[50]

Conference on Neural Information Processing Systems (NeurIPS) (2025)

Tesch, J., Becherini, G., Achar, P., Yiannakidis, A., Kocabas, M., Patel, P., Black, M.J.: BEDLAM2.0: Synthetic humans and cameras in motion. Conference on Neural Information Processing Systems (NeurIPS) (2025)

2025
[51]

In: International Conference on Computer Vision (ICCV) (2019) 18 OmniFit arXiv Preprint

Thomas, H., Qi, C.R., Deschaud, J.E., Marcotegui, B., Goulette, F., Guibas, L.J.: Kpconv: Flexible and deformable convolution for point clouds. In: International Conference on Computer Vision (ICCV) (2019) 18 OmniFit arXiv Preprint

2019
[52]

In: IEEE Conference Virtual Reality and 3D User Interfaces (VR) (2025)

Wang,B.,Meng,H.,Cao,R.,Cai,Z.,Li,L.,Ma,Y.,Chen,Q.,Wang,Z.:MagicScroll: Enhancing immersive storytelling with controllable scroll image generation. In: IEEE Conference Virtual Reality and 3D User Interfaces (VR) (2025)

2025
[53]

In: International Conference on 3D Vision (3DV) (2025)

Wang, D., Meng, H., Cai, Z., Shao, Z., Liu, Q., Wang, L., Fan, M., Zhan, X., Wang, Z.: Headevolver: Text to head avatars via expressive and attribute-preserving mesh deformation. In: International Conference on 3D Vision (3DV) (2025)

2025
[54]

arXiv preprint arXiv:2601.02267 (2025)

Wang, R., Zhang, Z., Tai, Y., Yang, J.: DiffProxy: Multi-view human mesh recovery via diffusion-generated dense proxies. arXiv preprint arXiv:2601.02267 (2025)

work page arXiv 2025
[55]

In: IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2021)

Wang, S., Geiger, A., Tang, S.: Locally aware piecewise transformation fields for 3d human mesh registration. In: IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2021)

2021
[56]

In: IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2024)

Wang, W., Ho, H.I., Guo, C., Rong, B., Grigorev, A., Song, J., Zarate, J.J., Hilliges, O.: 4d-dress: A 4d dataset of real-world human clothing with semantic annotations. In: IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2024)

2024
[57]

In: IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2025)

Wang, Y., Sun, Y., Patel, P., Daniilidis, K., Black, M.J., Kocabas, M.: Prompthmr: Promptable human mesh recovery. In: IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2025)

2025
[58]

In: IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2024)

Wu, X., Jiang, L., Wang, P.S., Liu, Z., Liu, X., Qiao, Y., Ouyang, W., He, T., Zhao, H.: Point transformer v3: simpler, faster, stronger. In: IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2024)

2024
[59]

Conference on Neural Information Processing Systems (NeurIPS) (2022)

Wu, X., Lao, Y., Jiang, L., Liu, X., Zhao, H.: Point transformer v2: grouped vector attention and partition-based pooling. Conference on Neural Information Processing Systems (NeurIPS) (2022)

2022
[60]

In: IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2025)

Xiang, J., Lv, Z., Xu, S., Deng, Y., Wang, R., Zhang, B., Chen, D., Tong, X., Yang, J.: Structured 3d latents for scalable and versatile 3d generation. In: IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2025)

2025
[61]

In: International Conference on Computer Vision (ICCV) (2023)

Yang, Z., Cai, Z., Mei, H., Liu, S., Chen, Z., Xiao, W., Wei, Y., Qing, Z., Wei, C., Dai, B., et al.: Synbody: Synthetic dataset with layered human models for 3d human perception and modeling. In: International Conference on Computer Vision (ICCV) (2023)

2023
[62]

In: IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2021)

Yu, T., Zheng, Z., Guo, K., Liu, P., Dai, Q., Liu, Y.: Function4d: Real-time human volumetric capture from very sparse consumer rgbd sensors. In: IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2021)

2021
[63]

In: IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2022)

Yu, X., Tang, L., Rao, Y., Huang, T., Zhou, J., Lu, J.: Point-bert: Pre-training 3d point cloud transformers with masked point modeling. In: IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2022)

2022
[64]

Conference on Neural Information Processing Systems (NeurIPS) (2017)

Zaheer, M., Kottur, S., Ravanbakhsh, S., Poczos, B., Salakhutdinov, R.R., Smola, A.J.: Deep sets. Conference on Neural Information Processing Systems (NeurIPS) (2017)

2017
[65]

In: IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2017)

Zhang, C., Pujades, S., Black, M.J., Pons-Moll, G.: Detailed, accurate, human shape estimation from clothed 3d scan sequences. In: IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2017)

2017
[66]

In: Interna- tional Conference on Computer Vision (ICCV) (2021)

Zhao, H., Jiang, L., Jia, J., Torr, P.H., Koltun, V.: Point transformer. In: Interna- tional Conference on Computer Vision (ICCV) (2021)

2021
[67]

Hunyuan3D 2.0: Scaling Diffusion Models for High Resolution Textured 3D Assets Generation

Zhao, Z., Lai, Z., Lin, Q., Zhao, Y., Liu, H., Yang, S., Feng, Y., Yang, M., Zhang, S., Yang, X., et al.: Hunyuan3d 2.0: Scaling diffusion models for high resolution textured 3d assets generation. arXiv preprint arXiv:2501.12202 (2025)

work page Pith review arXiv 2025
[68]

In: International Conference on Computer Vision (ICCV) (2019) OmniFit arXiv Preprint 19

Zheng, Z., Yu, T., Wei, Y., Dai, Q., Liu, Y.: Deephuman: 3d human reconstruction from a single image. In: International Conference on Computer Vision (ICCV) (2019) OmniFit arXiv Preprint 19

2019
[69]

Zuffi, S., Black, M.J.: The stitched puppet: A graphical model of 3d human shape and pose. In: IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2015) 20 OmniFit arXiv Preprint A Details of Unified Dataset To train our generalizable landmark predictor, we construct a unified dataset by combiningCAPE, 4D-DRESS, and two newly collected synt...

2015