DT-LSD: Deformable Transformer-based Line Segment Detection

TL;DR

This paper introduces DT-LSD, a deformable transformer-based model for line segment detection that is faster to train and more accurate than previous transformer and CNN-based methods, with significant improvements demonstrated on benchmark datasets.

Contribution

The paper proposes a novel deformable transformer model for line segment detection and a training acceleration technique called LCDN, outperforming existing models in speed and accuracy.

Findings

DT-LSD achieves higher accuracy than CNN-based models.

LCDN speeds up training by 34 times.

DT-LSD outperforms previous transformer-based models.

Abstract

Line segment detection is a fundamental low-level task in computer vision, and improvements in this task can impact more advanced methods that depend on it. Most new methods developed for line segment detection are based on Convolutional Neural Networks (CNNs). Our paper seeks to address challenges that prevent the wider adoption of transformer-based methods for line segment detection. More specifically, we introduce a new model called Deformable Transformer-based Line Segment Detection (DT-LSD) that supports cross-scale interactions and can be trained quickly. This work proposes a novel Deformable Transformer-based Line Segment Detector (DT-LSD) that addresses LETR's drawbacks. For faster training, we introduce Line Contrastive DeNoising (LCDN), a technique that stabilizes the one-to-one matching process and speeds up training by 34$\times$. We show that DT-LSD is faster and more accurate than its predecessor transformer-based model (LETR) and outperforms all CNN-based models in terms of accuracy. In the Wireframe dataset, DT-LSD achieves 71.7 for $sAP^{10}$ and 73.9 for $sAP^{15}$; while 33.2 for $sAP^{10}$ and 35.1 for $sAP^{15}$ in the YorkUrban dataset.