SMPL

By Matthew Maverick Loper, Naureen Mahmood, Javier Romero, Gerard Pons-Moll, Michael J. Black. ACM Transactions on Graphics, 2015

4,375 citations recorded by the source corpus.

Subfields: Computational Mechanics, Computer Graphics and Computer-Aided Design, Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/919b8e78179ef0364f66790e05c5e8712fcdd66b

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computational Mechanics

In Computer Graphics and Computer-Aided Design

In Computer Vision and Pattern Recognition

Citations in the corpus

Recent advances in 3D Gaussian splatting (2024, neutral): “As a special category, the human body can introduce the model prior, such as SMPL [172], to assist in generation.”
NeuMan: Neural Human Radiance Field from a Single Video (2022, positive): “Therefore, we define a canonical space based on the 大 -pose (Da-pose) SMPL [22] mesh, similar to [32,19].”
Nerfies: Deformable Neural Radiance Fields (2020, background): “Domain-Speciﬁc Modeling: Many reconstruction methods use domain-speciﬁc knowledge to model the shape and appearance of categories with limited topological variation, such as faces [4, 6, 8], human bodies [32, 61], and animals [13, 67].”
Fake it till you make it: face analysis in the wild using synthetic data alone (2021, background): “All clothing items were authored on an unclothed body mesh with either the average male or female body proportions [37] in a relaxed stance.”
Human Motion Diffusion as a Generative Prior (2023, methodology): “It is used as a prior for motion applications, such as motion denoising, fitting SMPL [Loper et al. 2015] model to joint location and as a pose code book for motion generation [Hong et al. 2022].”
BEHAVE: Dataset and Method for Tracking Human Object Interactions (2022, methodology): “Our dataset contains multi-view RGBD frames, with accurate pseudo-ground truth SMPL [52], object ﬁts, human and object segmentation masks, and contact annotations.”
Generating Diverse and Natural 3D Human Motions from Text (2022, methodology): “Motions from other comparison methods are too distorted to be rendered with the SMPL human shapes [22].”
Kimera: From SLAM to spatial perception with 3D dynamic scene graphs (2021, methodology): “The data-association check is made more robust by using the beta-parameters of the SMPL model (Loper et al., 2015), which encode the various shape attributes of the mesh in eight floating-point parameters.”
OpenPose: Realtime Multi-Person 2D Pose Estimation Using Part Affinity Fields (2018, background): “However, graphics applications such as avatar retargeting or 3D human shape reconstruction ([61], [68]) require foot keypoints such as big toe and heel.”
Perpetual Humanoid Control for Real-time Simulated Avatars (2023, methodology): “Our humanoid controller can support any human kinematic structure, and we use the SMPL [19] kinematic structure following prior arts [52, 20, 21].”
Image-Based 3D Object Reconstruction: State-of-the-Art and Trends in the Deep Learning Era (2019, methodology): “Parametric methods regularize the problem using statistical models such as morphable models [112], SCAPE [113], and SMPL [114].”
PaMIR: Parametric Model-Conditioned Implicit Representation for Image-Based Human Reconstruction (2020, methodology): “Inspired by DeepHuman [5], we integrate the parametric body model, SMPL [9], to regularize the human reconstruction.”
Action-Conditioned 3D Human Motion Synthesis with Transformer VAE (2021, methodology): “Consequently, without loss of generality, we employ the SMPL body model [46], which is a disentangled body representation (similar to recent models [53, 55, 58, 68]).”
Blending Anti-Aliasing into Vision Transformer (2021, methodology): “The linear assembling strategy has been proven effective in computational photography [43], parametric human body representation [44], network compression [45] and acceleration [46].”
Champ: Controllable and Consistent Human Image Animation with 3D Parametric Guidance (2024, methodology): “The SMPL [20] model is a 3D model renowned for its realistic depiction of human bodies through skinning and blend shapes.”