ScaleLSD: Scalable Deep Line Segment Detection Streamlined

2025-06-11

AI Summary

This paper addresses domain-agnostic, robust line segment detection in natural images. Methodologically, it introduces the first scalable, self-supervised, zero-shot generalizable deep line detector: a lightweight CNN architecture that fuses multi-scale features and is trained at scale—over ten million unlabeled real-world images—via self-supervision. A unified zero-shot transfer evaluation framework is proposed to assess generalization across diverse geometric vision tasks, including single-/dual-/multi-view correspondence and 3D line mapping. Experimentally, the model achieves state-of-the-art performance—outperforming the classical LSD algorithm across all standard detection metrics for the first time—with substantial gains in segment completeness and geometric accuracy. Moreover, it demonstrates strong zero-shot transfer capability without task-specific fine-tuning. This work establishes a general, efficient, and annotation-free paradigm for image geometric modeling.

Technology Category

• Computer Vision: Segmentation
• Machine Learning: Unsupervised & Self-Supervised Learning
• Search and Optimization: Learning to Search

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Semantics and Knowledge: Methods, algorithms and applications for the development of semantic models, knowledge graphs and other forms of structured data models with machine-interpretable semantics
• Web Mining and Content Analysis: Robustness and generalizability of Web mining methods

Abstract

This paper studies the problem of Line Segment Detection (LSD) for the characterization of line geometry in images, with the aim of learning a domain-agnostic robust LSD model that works well for any natural images. With the focus of scalable self-supervised learning of LSD, we revisit and streamline the fundamental designs of (deep and non-deep) LSD approaches to have a high-performing and efficient LSD learner, dubbed as ScaleLSD, for the curation of line geometry at scale from over 10M unlabeled real-world images. Our ScaleLSD works very well to detect much more number of line segments from any natural images even than the pioneered non-deep LSD approach, having a more complete and accurate geometric characterization of images using line segments. Experimentally, our proposed ScaleLSD is comprehensively testified under zero-shot protocols in detection performance, single-view 3D geometry estimation, two-view line segment matching, and multiview 3D line mapping, all with excellent performance obtained. Based on the thorough evaluation, our ScaleLSD is observed to be the first deep approach that outperforms the pioneered non-deep LSD in all aspects we have tested, significantly expanding and reinforcing the versatility of the line geometry of images. Code and Models are available at https://github.com/ant-research/scalelsd

Problem

• Develops domain-agnostic robust Line Segment Detection model
• Streamlines LSD designs for scalable self-supervised learning
• Enhances line geometry characterization in natural images

Innovation

• Scalable self-supervised learning for LSD
• Streamlined deep and non-deep LSD designs
• High-performing efficient learner for line geometry