Skip to content

Code for the paper "ANISE: Assembly-based Neural Implicit Surface rEconstruction"

Notifications You must be signed in to change notification settings

lodurality/anise_paper_code

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

3 Commits
 
 

Repository files navigation

ANISE: Assembly-based Neural Implicit Surface rEconstruction

This is the official code for ANISE, an efficient neural part-aware shape representation that models 3D shapes as collection of neural part implicits that are assembled in the full shape. This representation allows for editable part-aware single-view and point cloud reconstruction and reconstruction conditioned on small set of reference parts. We establish current state-of-the-art in part-aware single view reconstruction and reconstruction based on part-retrieval.

ANISE: Assembly-based Neural Implicit Surface rEconstruction

Project Page | Paper (TVCG)

ANISE pipeline

An overview of our approach. ANISE consists of three modules. The first module predicts a coarse part arrangement in the form of part transformations (structure prediction). Then, conditioned on these part transformations, the next module predicts an implicit function representing the actual geometry (geometry prediction). The last module transforms each part implicit functions according to the predicted transformations and combines them into a single output implicit function (assembly).

About

Code for the paper "ANISE: Assembly-based Neural Implicit Surface rEconstruction"

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published