SCAPE is a data-driven method for building a human shape model that spans variation in both subject shape and pose. The method incorporates both articulated and non-rigid deformations. It learns a pose deformation model that derives non-rigid surface deformation as a function of the articulated skeleton's pose, and a separate model of body shape variation. These models can be combined to produce 3D surface models with realistic muscle deformation for different people in different poses, even when neither appear in the training set. The method is used for shape completion, generating complete surface meshes from limited marker data. It also enables motion capture animation using a single static scan and marker data. The model allows for variation in individual body shape and can synthesize animations with realistic muscle deformation. The SCAPE model decouples pose and body shape deformation, simplifying the mathematical formulation and enabling efficient learning. However, it cannot capture strong correlations between body shape and muscle deformation. The model is focused on muscle deformations from articulated motion and does not encode deformations from other factors. It is data-driven, generating models from scans, with minimal human intervention for marker placement. The framework can be applied to other data sets and is generalizable to cases involving articulated motion. The work was supported by grants from ONR and acknowledges the help of collaborators in providing scan data.SCAPE is a data-driven method for building a human shape model that spans variation in both subject shape and pose. The method incorporates both articulated and non-rigid deformations. It learns a pose deformation model that derives non-rigid surface deformation as a function of the articulated skeleton's pose, and a separate model of body shape variation. These models can be combined to produce 3D surface models with realistic muscle deformation for different people in different poses, even when neither appear in the training set. The method is used for shape completion, generating complete surface meshes from limited marker data. It also enables motion capture animation using a single static scan and marker data. The model allows for variation in individual body shape and can synthesize animations with realistic muscle deformation. The SCAPE model decouples pose and body shape deformation, simplifying the mathematical formulation and enabling efficient learning. However, it cannot capture strong correlations between body shape and muscle deformation. The model is focused on muscle deformations from articulated motion and does not encode deformations from other factors. It is data-driven, generating models from scans, with minimal human intervention for marker placement. The framework can be applied to other data sets and is generalizable to cases involving articulated motion. The work was supported by grants from ONR and acknowledges the help of collaborators in providing scan data.