MoSh

By Matthew Maverick Loper, Naureen Mahmood, Michael J. Black. 2014

235 citations recorded by the source corpus.

Subfields: Computational Mechanics, Computer Vision and Pattern Recognition, Control and Systems Engineering

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 Vision and Pattern Recognition

In Control and Systems Engineering

Citations in the corpus

End-to-End Recovery of Human Shape and Pose (2017, methodology): “We thank Naureen Mahmood for providing MoShed datasets and mesh retargeting for character animation, Dushyant Mehta for his assistance on MPIINF-3DHP, and Shubham Tulsiani, Abhishek Kar, Saurabh Gupta, David Fouhey and Ziwei Liu for helpful discussions.”
PyMAF: 3D Human Pose and Shape Regression with Pyramidal Mesh Alignment Feedback Loop (2021, methodology): “6M are generated by applying MoSh [12] to the sparse 3D Mo-Cap marker data, as done in Kanazawa et al .”
Unite the People: Closing the Loop Between 3D and 2D Human Representations (2017, methodology): “The improved predictions from the 91 landmark model increase the ratio of high quality 3D fits on the LSP dataset by 9.3% when compared to the fits using 14 keypoint ground truth locations (Sec.”
GOAL: Generating 4D Whole-Body Motion for Hand-Object Grasping (2021, background): “These observations are in line with recent work [36,62,64], but we go beyond them in regressing 3D offsets together with SMPL-X parameters, instead of regressing point positions and ﬁtting SMPL-X to these.”
Neural Body Fitting: Unifying Deep Learning and Model Based Human Pose and Shape Estimation (2018, methodology): “However, by applying MoSH [29] to the 3D MoCap marker data provided by the latter we obtain the corresponding SMPL parameters, which in turn allows us to generate part labels by rendering an appropriately annotated SMPL mesh [24].”
Expressive Body Capture: 3D Hands, Face, and Body From a Single Image (2019, unknown): “These parameters were originally released by [19, 21].”
DeepHuman: 3D Human Reconstruction From a Single Image (2019, methodology): “In order to obtain human mesh data under natural but diverse poses, our system presents to the subject a reference pose randomly sampled from MOSH[40] dataset every 6 seconds and requires the performer to imitate the reference pose in the second stage.”
STAR: Sparse Trained Articulated Human Body Regressor (2020, methodology): “f pose . The STAR model is fully parameterized by 72 (i.e. 24 * 3) pose parameters in axis-angle representation, and up to 300 shape parameters . 3.1 Model Training STAR training is similar to SMPL [22]. The key dierence is the training of the pose corrective function in Equation 3. STAR pose correc”
Learning from Synthetic Humans (2017, methodology): “For this reason we do not use the skeleton data but rather use MoSh [20] to fit the SMPL parameters that best explain raw 3D MoCap marker locations.”
Learning a Deep Model for Human Action Recognition from Novel Viewpoints (2016, methodology): “Several approaches [59], [60] have been proposed in the literatures to”
3D Action Recognition from Novel Viewpoints (2016, methodology): “Several methods [4, 30] have been proposed to fit a human model to motion capture skeleton data of a person.”
Keep It SMPL: Automatic Estimation of 3D Human Pose and Shape from a Single Image (2016, methodology): “We train a prior over pose from SMPL models that have been ﬁt to the CMU mocap marker data [3] using MoSh [29].”
I2L-MeshNet: Image-to-Lixel Prediction Network for Accurate 3D Human Pose and Mesh Estimation from a Single RGB Image (2020, methodology): “All the previous works [20, 23, 24, 40] used SMPL parameters obtained by applying Mosh [27] on the marker data of Human3.”
MotionBERT: A Unified Perspective on Learning Human Motion Representations (2022, methodology): “AMASS [71] integrates most existing marker-based Mocap datasets [1,2,4,10,14,17,32,35,48,65,67,72,79,90,103,104] and parameterizes them with a common representation.”
Pose2Mesh: Graph Convolutional Network for 3D Human Pose and Mesh Recovery from a 2D Human Pose (2020, methodology): “ture system, but there are no groundtruth 3D meshes. As a result, for 3D mesh supervision, most of the previous 3D pose and mesh estimation works [26,30,31] used pseudo-groundtruth obtained from Mosh [35]. However, because of the license issue, the pseudo-groundtruth from Mosh is not currently publi”