computer-vision

3D Multi-Person Mesh Recovery

3260 papers • 126 benchmarks • 313 datasets

This task has no description! Would you like to contribute one?

(Image credit: Papersgraph)

Benchmarks

These leaderboards are used to track progress in 3d-multi-person-mesh-recovery

Trend

Dataset

Best Model

Actions

AGORA

Relative Human

Libraries

i

Use these libraries to find 3d-multi-person-mesh-recovery models and implementations

Datasets

AGORA

Relative Human

Subtasks

No subtasks available.

Most implemented papers

Monocular, One-stage, Regression of Multiple 3D People

Michael J. Black, Tao Mei, Yu Sun, Qian Bao, Yili Fu, Wu Liu•Wed Aug 26 2020

ROMP is the first real-time implementation of monocular multi-person 3D mesh regression, and achieves superior performance on the challenging multi- person benchmarks, including 3DPW and CMU Panoptic.

321

Content

0

Paper Graph

Monocular Expressive Body Regression through Body-Driven Attention

Michael J. Black, G. Pavlakos, Timo Bolkart, Dimitrios Tzionas, Vasileios Choutas•Wed Aug 19 2020

ExPose estimates expressive 3D humans more accurately than existing optimization methods at a small fraction of the computational cost by introducing ExPose (EXpressive POse and Shape rEgression), which directly regresses the body, face, and hands, in SMPL-X format, from an RGB image.

270 0

Paper Graph

AGORA: Avatars in Geography Optimized for Regression Analysis

Michael J. Black, Priyanka Patel, C. Huang, J. Tesch, David T. Hoffmann, Shashank Tripathi•Wed Apr 28 2021

This work introduces AGORA, a synthetic dataset with high realism and highly accurate ground truth, and evaluates existing state-of-the-art methods for 3D human pose estimation on this dataset, finding that most methods perform poorly on images of children.

257 0

Paper Graph

Expressive Body Capture: 3D Hands, Face, and Body From a Single Image

Ahmed A. A. Osman, Michael J. Black, G. Pavlakos, Timo Bolkart, Dimitrios Tzionas, N. Ghorbani, Vasileios Choutas•Wed Apr 10 2019

This work uses the new method, SMPLify-X, to fit SMPL-X to both controlled images and images in the wild, and evaluates 3D accuracy on a new curated dataset comprising 100 images with pseudo ground-truth.

2139 0

Paper Graph

Body Size and Depth Disambiguation in Multi-Person Reconstruction from Single Images

F. Moreno-Noguer, A. Sanfeliu, Nicolas Ugrinovic, Adria Ruiz, Antonio Agudo•Mon Nov 01 2021

This work devises a novel optimization scheme that learns the appropriate body scale and relative camera pose, by enforcing the feet of all people to remain on the ground floor, and is able to robustly estimate the body translation and shape of multiple people while retrieving their spatial arrangement.

11 0

Paper Graph

One-Stage 3D Whole-Body Mesh Recovery with Component Aware Transformer

Jing-de Lin, Ailing Zeng, Haoqian Wang, Lei Zhang, Y. Li•Mon Mar 27 2023

This work designs a one-stage pipeline for expressive whole-body mesh recovery, named OSX, without separate networks for each part, and designs a Component Aware Transformer (CAT) composed of a global body encoder and a local face/hand decoder.

146 0

Paper Graph

Reconstructing Groups of People with Hypergraph Relational Reasoning

Buzhen Huang, Jingyi Ju, Zhihao Li, Yangang Wang•Tue Aug 29 2023

A novel hypergraph relational reasoning network is proposed to formulate the complex and high-order relation correlations among individuals and groups in the crowd to produce accurate absolute body poses and shapes in large-scale crowded scenes.

19 0

Paper Graph

SMPLer-X: Scaling Up Expressive Human Pose and Shape Estimation

Chen Change Loy, Zhongang Cai, Lei Yang, Wanqi Yin, Ailing Zeng, Ziwei Liu, Lei Zhang, Chen Wei, Qingping Sun, Yanjun Wang, Hui En Pang, Haiyi Mei, Mingyuan Zhang•Thu Sep 28 2023

This work investigates scaling up EHPS towards the first generalist foundation model (dubbed SMPLer-X), with up to ViT-Huge as the backbone and training with upto 4.5M instances from diverse data sources.

151 0

Paper Graph

Multi-HMR: Multi-Person Whole-Body Human Mesh Recovery in a Single Shot

Philippe Weinzaepfel, Grégory Rogez, Thomas Lucas, Fabien Baradel, M. Armando, Salma Galaaoui, Romain Br'egier•Wed Feb 21 2024

This work presents Multi-HMR, a strong sigle-shot model for multi-person 3D human mesh recovery from a single RGB image, and introduces CUFFS, the Close-Up Frames of Full-Body Subjects dataset, containing humans close to the camera with diverse hand poses.

66 0

Paper Graph

Collaborative Regression of Expressive Bodies using Moderation

Michael J. Black, Timo Bolkart, Dimitrios Tzionas, Vasileios Choutas, Yao Feng•Mon May 10 2021

PIXIE is introduced, which produces animatable, whole-body 3D avatars with realistic facial detail, from a single image and is shown to be more accurate whole-shape and detailed face shape than the state of the art.

178 0

Paper Graph

Adding a benchmark result helps the community track progress.