Keep It SMPL: Automatic Estimation of 3D Human Pose and Shape from a Single Image

@inproceedings{Bogo2016KeepIS,
  title={Keep It SMPL: Automatic Estimation of 3D Human Pose and Shape from a Single Image},
  author={Federica Bogo and Angjoo Kanazawa and Christoph Lassner and Peter V. Gehler and Javier Romero and Michael J. Black},
  booktitle={ECCV},
  year={2016}
}
We describe the first method to automatically estimate the 3D pose of the human body as well as its 3D shape from a single unconstrained image. [...] Key Method To solve this, we first use a recently published CNN-based method, DeepCut, to predict (bottom-up) the 2D body joint locations. We then fit (top-down) a recently published statistical body shape model, called SMPL, to the 2D joints.Expand
DenseBody: Directly Regressing Dense 3D Human Pose and Shape From a Single Color Image
TLDR
An efficient representation of 3D human shape and pose which can be predicted through an encoder-decoder neural network is used which achieves state-of-the-art performance on several3D human body datasets including Human3.6M, SURREAL and UP-3D with even faster running speed. Expand
Learning to Estimate 3D Human Pose and Shape from a Single Color Image
TLDR
This work addresses the problem of estimating the full body 3D human pose and shape from a single color image and proposes an efficient and effective direct prediction method based on ConvNets, incorporating a parametric statistical body shape model (SMPL) within an end-to-end framework. Expand
SMPLR: Deep SMPL reverse for 3D human pose and shape recovery
TLDR
This paper proposes to embed SMPL within a deep model to accurately estimate 3D pose and shape from a still RGB image, and proposes a denoising autoencoder (DAE) module between CNN and SMPLR, able to lift 2D joints to 3D and partially recover from structured error. Expand
Template based Human Pose and Shape Estimation from a Single RGB-D Image
TLDR
This work proposes to utilize both 2D and 3D joint points, which provide more information, to fit a parametric body model (SMPL) to reconstruct the 3D human body model from a single RGB-D image. Expand
Unite the People: Closing the Loop Between 3D and 2D Human Representations
TLDR
This work proposes a hybrid approach to 3D body model fits for multiple human pose datasets with an extended version of the recently introduced SMPLify method, and shows that UP-3D can be enhanced with these improved fits to grow in quantity and quality, which makes the system deployable on large scale. Expand
3D Human Pose Estimation from a Single Image via Distance Matrix Regression
  • F. Moreno-Noguer
  • Computer Science
  • 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)
  • 2017
TLDR
It is shown that more precise pose estimates can be obtained by representing both the 2D and 3D human poses using NxN distance matrices, and formulating the problem as a 2D-to-3D distance matrix regression. Expand
Beyond Weak Perspective for Monocular 3D Human Pose Estimation
TLDR
This algorithm has resulted in a winning entry for the 3DPW Challenge, reaching first place in joints orientation accuracy and using full perspective projection, with the correct camera center and an approximated focal length, provides favorable results. Expand
3D Body Pose and Shape Estimation from Multi-View Images With Limb Geometric Constraint
TLDR
This paper brings up an end-to-end method to estimate 3D human shape and pose from multi-view RGB images and employs the idea of adversarial learning in the model to help regress accurate pose and shape parameters. Expand
BodyNet: Volumetric Inference of 3D Human Body Shapes
TLDR
BodyNet is an end-to-end trainable network that benefits from a volumetric 3D loss, a multi-view re-projection loss, and intermediate supervision of 2D pose, 2D body part segmentation, and 3D pose and achieves state-of-the-art performance. Expand
SMPLy Benchmarking 3D Human Pose Estimation in the Wild
TLDR
A pipeline to easily produce and validate such a dataset with accurate ground-truth, with which to benchmark recent 3D human pose estimation methods in-the-wild on the recently introduced Mannequin Challenge dataset is presented. Expand
...
1
2
3
4
5
...

References

SHOWING 1-10 OF 59 REFERENCES
Estimating human shape and pose from a single image
TLDR
A solution to the challenging problem of estimating human body shape from a single photograph or painting is described and a novel application in which 3D human models are extracted from archival photographs and paintings is demonstrated. Expand
Robust Estimation of 3D Human Poses from a Single Image
TLDR
This work proposes a method of estimating 3D human poses from a single image, which works in conjunction with an existing 2D pose/joint detector, which outperforms the state-of-the-arts on three benchmark datasets. Expand
A Joint Model for 2D and 3D Pose Estimation from a Single Image
TLDR
A Bayesian framework is proposed that integrates a generative model based on latent variables and discriminative 2D part detectors based on HOGs, and performs inference using evolutionary algorithms that provide accurate 2D and 3D pose estimations even when the 2D detectors are inaccurate. Expand
Sparseness Meets Deepness: 3D Human Pose Estimation from Monocular Video
TLDR
This paper addresses the challenge of 3D full-body human pose estimation from a monocular image sequence with a novel approach that integrates a sparsity-driven 3D geometric prior and temporal smoothness and outperforms a publicly available 2D pose estimation baseline on the challenging PennAction dataset. Expand
Pose-conditioned joint angle limits for 3D human pose reconstruction
TLDR
A general parametrization of body pose is defined and a new, multi-stage, method to estimate 3D pose from 2D joint locations using an over-complete dictionary of poses is defined that shows good generalization while avoiding impossible poses. Expand
Iterated Second-Order Label Sensitive Pooling for 3D Human Pose Estimation
TLDR
This paper provides evidence for a positive answer for 3D pose estimation from RGB images by leveraging 2D human body part labeling in images, second-order label-sensitive pooling over dynamically computed regions resulting from a hierarchical decomposition of the body, and iterative structured-output modeling to contextualize the process based on3D pose estimates. Expand
Detailed Human Shape and Pose from Images
TLDR
This work represents the body using a recently proposed triangulated mesh model called SCAPE which employs a low-dimensional, but detailed, parametric model of shape and pose-dependent deformations that is learned from a database of range scans of human bodies. Expand
3D Human Pose Reconstruction Using Millions of Exemplars
  • Hao Jiang
  • Computer Science
  • 2010 20th International Conference on Pattern Recognition
  • 2010
TLDR
This work proposes a novel exemplar based method to estimate 3D human poses from single images by using only the joint correspondences, which implicitly squares the size of the exemplar database and enables us to reconstruct unconstrained poses efficiently. Expand
Depth Sweep Regression Forests for Estimating 3D Human Pose from Images
TLDR
This work extends regression forests to infer missing depth data of image features and 3D pose simultaneously, and hypothesizes the depth of the features by sweeping with a plane through the 3D volume of potential joint locations. Expand
Reconstructing 3D Human Pose from 2D Image Landmarks
TLDR
This work presents an activity-independent method to recover the 3D configuration of a human figure from 2D locations of anatomical landmarks in a single image, leveraging a large motion capture corpus as a proxy for visual memory. Expand
...
1
2
3
4
5
...