SCAPE: shape completion and animation of people

By Dragomir Anguelov, Praveen Srinivasan, Daphne Koller, Sebastian Thrun, Jim Rodgers, James Davis. 2005

1,984 citations recorded by the source corpus.

We introduce the SCAPE method (Shape Completion and Animation for PEople)---a data-driven method for building a human shape model that spans variation in both subject shape and pose. The method is based on a representation that incorporates both articulated and non-rigid deformations. We learn a pose deformation model that derives the non-rigid surface deformation as a function of the pose of the articulated skeleton. We also learn a separate model of variation based on body shape. Our two models can be combined to produce 3D surface models with realistic muscle deformation for different people in different poses, when neither appear in the training set. We show how the model can be used for shape completion --- generating a complete surface mesh given a limited set of markers specifying the target shape. We present applications of shape completion to partial view completion and motion capture animation. In particular, our method is capable of constructing a high-quality animated surface model of a moving person, with realistic muscle deformation, using just a single static scan and a marker motion capture sequence of the person.

Subfields: 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 Computer Vision and Pattern Recognition

In Control and Systems Engineering

Citations in the corpus

Geometric Deep Learning on Graphs and Manifolds Using Mixture Model CNNs (2016, methodology): “We also show correspondence on shapes from SCAPE [2] and TOSCA [9] datasets.”
Joint Training of a Convolutional Network and a Graphical Model for Human Pose Estimation (2014, neutral): “More complex examples come from the HumanEva dataset competitions [33] that use video or higher-resolution shape models such as SCAPE [3] and extensions.”
Learning Category-Specific Mesh Reconstruction from Image Collections (2018, methodology): “These techniques have since been applied to model human bodies [1,19], hands [27,17], and more recently on quadruped animals [40].”
Lifting Object Detection Datasets into 3D (2015, neutral): “To account for differences between the 3D exemplars and the object depicted, Su and Guibas [26] choose a few exemplars that appear to be similar to the depicted shape and combine them to produce a single depth map.”
Category-specific object reconstruction from a single image (2014, neutral): “Here we pursue more expressive basis shape models [1, 7, 42] which establish a balance between the two extremes as they can deform but only along class-specific modes of variation.”
Learning Gradient Fields for Shape Generation (2020, neutral): “Common representations emerged for deep generative 3D modeling include voxel-based [23,63], mesh-based [3,48,20,27,35,57], and assembly-based techniques [34,41].”
Dense Human Body Correspondences Using Convolutional Networks (2015, methodology): “Note that the Yobi3D dataset covers the shape variability in local geometry, while the SCAPE and the MIT datasets cover the variability in pose.”
Deep Functional Maps: Structured Prediction for Dense Shape Correspondence (2017, methodology): “We demonstrate the generalization capabilities of FMNet by performing a series of experiments on the SCAPE dataset of human shapes [4], where our network is trained on FAUST scans data as described previously.”
Learning Human Pose Estimation Features with Convolutional Networks (2013, methodology): “Further examples come from the HumanEva dataset competitions [41], or approaches that use higher-resolution shape models such as SCAPE [4] and further extensions [20, 8].”
Deformable Shape Completion with Graph Convolutional Autoencoders (2017, methodology): “This distinction becomes crucial as large missing regions force the priors to become more complex (see for example the human model designed in [2]).”
Monocular Total Capture: Posing Face, Body, and Hands in the Wild (2018, methodology): “3D Deformable Human Models: 3D deformable models are commonly used for markerless body [6, 30, 43] and face motion capture [8, 13] to restrict the reconstruction output to the shape and motion spaces defined by the models.”
What Shape Are Dolphins? Building 3D Morphable Models from 2D Images (2013, neutral): “ORPHABLE models of the human head and body have enabled a myriad of applications in computer graphics, special effects, and computational photography [1], [2], [3], [4], [5].”
Deformation-driven shape correspondence via shape recognition (2017, background): “Early works on capturing statistical variations of the human body explored smooth deformations of a ﬁxed template [Blanz and Vetter 1999; Allen et al. 2003; Anguelov et al. 2005].”
Learning class‐specific descriptors for deformable shapes using localized spectral convolutional networks (2015, methodology): “Datasets We used two public-domain datasets of scanned human shapes in different poses: SCAPE [A ∗ 05] and FAUST [BRLB14], the latter being the most recent and particularly challenging, given a high variability of non-isometric deformations as well as signiﬁcant variability between different human…”
Learning a model of facial shape and expression from 4D scans (2017, methodology): “Our FLAME model (Faces Learned with an Articulated Model and Expressions) is factored in that it separates the representation of identity, pose, and facial expression, similar to models of the human body [Anguelov et al. 2005; Loper et al. 2015].”
Perceiving 3D Human-Object Spatial Arrangements from a Single Image in the Wild (2020, methodology): “As such, most methods employ statistical 3D body models with strong priors on shape learned from large-scale 3D scans and with known kinematic structure to model the articulation [5,69,43,49,29].”
STAR: Sparse Trained Articulated Human Body Regressor (2020, methodology): “iterature on 3D modelling of the human body, either manually or using data-driven methods. We review the most related literature here with a focus on methods that learn bodies from data, pioneered by [4,6]. Linear Blend Skinning. Linear Blend Skinning (LBS), also known as SkeletalSubspace Deformatio”