SimpleProc: Fully Procedural Synthetic Data from Simple Rules for Multi-View Stereo

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Ma, Zeyu, Raistrick, Alexander, Deng, Jia
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910109952638976
author Ma, Zeyu
Raistrick, Alexander
Deng, Jia
author_facet Ma, Zeyu
Raistrick, Alexander
Deng, Jia
contents In this paper, we explore the design space of procedural rules for multi-view stereo (MVS). We demonstrate that we can generate effective training data using SimpleProc: a new, fully procedural generator driven by a very small set of rules using Non-Uniform Rational Basis Splines (NURBS), as well as basic displacement and texture patterns. At a modest scale of 8,000 images, our approach achieves superior results compared to manually curated images (at the same scale) sourced from games and real-world objects. When scaled to 352,000 images, our method yields performance comparable to--and in several benchmarks, exceeding--models trained on over 692,000 manually curated images. The source code and the data are available at https://github.com/princeton-vl/SimpleProc.
format Preprint
id arxiv_https___arxiv_org_abs_2604_04925
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle SimpleProc: Fully Procedural Synthetic Data from Simple Rules for Multi-View Stereo
Ma, Zeyu
Raistrick, Alexander
Deng, Jia
Computer Vision and Pattern Recognition
In this paper, we explore the design space of procedural rules for multi-view stereo (MVS). We demonstrate that we can generate effective training data using SimpleProc: a new, fully procedural generator driven by a very small set of rules using Non-Uniform Rational Basis Splines (NURBS), as well as basic displacement and texture patterns. At a modest scale of 8,000 images, our approach achieves superior results compared to manually curated images (at the same scale) sourced from games and real-world objects. When scaled to 352,000 images, our method yields performance comparable to--and in several benchmarks, exceeding--models trained on over 692,000 manually curated images. The source code and the data are available at https://github.com/princeton-vl/SimpleProc.
title SimpleProc: Fully Procedural Synthetic Data from Simple Rules for Multi-View Stereo
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2604.04925