POC-SLT: Partial Object Completion with SDF Latent Transformers

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Zakeri, Faezeh, Braun, Raphael, Ruppert, Lukas, Lensch, Henrik P. A.
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866929583712894976
author Zakeri, Faezeh
Braun, Raphael
Ruppert, Lukas
Lensch, Henrik P. A.
author_facet Zakeri, Faezeh
Braun, Raphael
Ruppert, Lukas
Lensch, Henrik P. A.
contents 3D geometric shape completion hinges on representation learning and a deep understanding of geometric data. Without profound insights into the three-dimensional nature of the data, this task remains unattainable. Our work addresses this challenge of 3D shape completion given partial observations by proposing a transformer operating on the latent space representing Signed Distance Fields (SDFs). Instead of a monolithic volume, the SDF of an object is partitioned into smaller high-resolution patches leading to a sequence of latent codes. The approach relies on a smooth latent space encoding learned via a variational autoencoder (VAE), trained on millions of 3D patches. We employ an efficient masked autoencoder transformer to complete partial sequences into comprehensive shapes in latent space. Our approach is extensively evaluated on partial observations from ShapeNet and the ABC dataset where only fractions of the objects are given. The proposed POC-SLT architecture compares favorably with several baseline state-of-the-art methods, demonstrating a significant improvement in 3D shape completion, both qualitatively and quantitatively.
format Preprint
id arxiv_https___arxiv_org_abs_2411_05419
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle POC-SLT: Partial Object Completion with SDF Latent Transformers
Zakeri, Faezeh
Braun, Raphael
Ruppert, Lukas
Lensch, Henrik P. A.
Computer Vision and Pattern Recognition
I.4.8; I.4.5
3D geometric shape completion hinges on representation learning and a deep understanding of geometric data. Without profound insights into the three-dimensional nature of the data, this task remains unattainable. Our work addresses this challenge of 3D shape completion given partial observations by proposing a transformer operating on the latent space representing Signed Distance Fields (SDFs). Instead of a monolithic volume, the SDF of an object is partitioned into smaller high-resolution patches leading to a sequence of latent codes. The approach relies on a smooth latent space encoding learned via a variational autoencoder (VAE), trained on millions of 3D patches. We employ an efficient masked autoencoder transformer to complete partial sequences into comprehensive shapes in latent space. Our approach is extensively evaluated on partial observations from ShapeNet and the ABC dataset where only fractions of the objects are given. The proposed POC-SLT architecture compares favorably with several baseline state-of-the-art methods, demonstrating a significant improvement in 3D shape completion, both qualitatively and quantitatively.
title POC-SLT: Partial Object Completion with SDF Latent Transformers
topic Computer Vision and Pattern Recognition
I.4.8; I.4.5
url https://arxiv.org/abs/2411.05419