Holistic Geometric Feature Learning for Structured Reconstruction

TL;DR

This paper introduces F-Learn, a frequency-domain feature learning strategy that enhances structured reconstruction by holistically integrating geometric information, leading to improved topology inference and primitive detection.

Contribution

The paper proposes a novel frequency-domain approach for holistic geometric feature learning, improving topological reasoning in structured reconstruction tasks.

Findings

F-Learn effectively incorporates structure awareness into geometric primitive detection.

Significant performance improvements in topology inference and structured reconstruction.

Lightweight design allows easy integration into existing deep models.

Abstract

The inference of topological principles is a key problem in structured reconstruction. We observe that wrongly predicted topological relationships are often incurred by the lack of holistic geometry clues in low-level features. Inspired by the fact that massive signals can be compactly described with frequency analysis, we experimentally explore the efficiency and tendency of learning structure geometry in the frequency domain. Accordingly, we propose a frequency-domain feature learning strategy (F-Learn) to fuse scattered geometric fragments holistically for topology-intact structure reasoning. Benefiting from the parsimonious design, the F-Learn strategy can be easily deployed into a deep reconstructor with a lightweight model modification. Experiments demonstrate that the F-Learn strategy can effectively introduce structure awareness into geometric primitive detection and topology inference, bringing significant performance improvement to final structured reconstruction. Code and pre-trained models are available at https://github.com/Geo-Tell/F-Learn.