shunsukesaito / SCANimate

This repository contains the code for the paper "SCANimate: Weakly Supervised Learning of Skinned Clothed Avatar Networks"
https://scanimate.is.tue.mpg.de/
Other
274 stars 31 forks source link

question: is there a recommended way or tool to get the pose and shape parameters of the input 3D scans #5

Closed codesavory closed 2 years ago

codesavory commented 3 years ago
  1. How to fit a SMPL model to the input 3D scan/mesh - how to extract the pose and shape parameters of the 3D scans is this done along-side scan generation or can it be done posterior to any arbitrary 3D mesh/3d scan
  2. Do you intend on providing any tools/suggest methods to extract the SMPL minimal and pose parameters for a given 3D mesh
JinlongYANG commented 3 years ago
  1. How to fit a SMPL model to the input 3D scan/mesh - how to extract the pose and shape parameters of the 3D scans is this done along-side scan generation or can it be done posterior to any arbitrary 3D mesh/3d scan

Generally speaking, the body pose is not explicitly captured in most commercial body scanners. So for most of the cases (scans from the CAPE dataset are in this situation), body pose retrieval is independent post-processing of capturing 3D scans.

  1. Do you intend on providing any tools/suggest methods to extract the SMPL minimal and pose parameters for a given 3D mesh

Unfortunately, we are not planning to release such a tool but there are many candidates. For example: a. One can following the work of ClothCap to retrieve body pose and shape while registering the surface. b. Or one can use Openpose result from multi-view images of the scans as initialization for the human pose, then run some optimization to obtain the best body fitting, as described in AGORA