arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2106.00186cs.CVcs.LG
收藏

一种轻量级的实时直线段检测方法

Towards Light-weight and Real-time Line Segment Detection

NAVER/LINE Vision(NAVER/LINE 视觉)

Geonmo Gu, Byungsoo Ko, SeoungHyun Go, Sung-Hyun Lee, Jingeun Lee, Minchul Shin

摘要:

先前基于深度学习的线段检测(LSD)遭受用于线预测的巨大模型大小和高计算成本。这限制了它们在计算受限的环境中进行实时的【形】推理。在本文中,我们提出了一个实时的【形】和轻量级的线段检测器资源受限的环境中命名为移动LSD(M-LSD)。我们设计了一个非常有效的LSD架构,最大限度地减少骨干网络,并删除了典型的多模块过程中发现以前的方法线预测。为了保持轻量级网络的竞争力,我们提出了新的训练方案:线段(SoL)增强,匹配和几何损失。SoL增强将线段分割成多个子部分,这些子部分用于在训练过程中提供辅助线数据。此外,匹配和几何损失允许模型捕获额外的几何线索。与之前最好的实时的【形】LSD方法TP-LSD-Lite相比,我们的模型(M-LSD-tiny)以2.5%的模型大小和130.5%的GPU推理速度获得了具有竞争力的性能。此外,我们的模型在最新的Android和iPhone移动设备上分别以56.8 FPS和48.6 FPS运行。据我们所知,这是移动设备上第一个实时的【形】深度LSD。我们的代码可用。

英文摘要:

Previous deep learning-based line segment detection (LSD) suffers from the immense model size and high computational cost for line prediction. This constrains them from real-time inference on computationally restricted environments. In this paper, we propose a real-time and light-weight line segment detector for resource-constrained environments named Mobile LSD (M-LSD). We design an extremely efficient LSD architecture by minimizing the backbone network and removing the typical multi-module process for line prediction found in previous methods. To maintain competitive performance with a light-weight network, we present novel training schemes: Segments of Line segment (SoL) augmentation, matching and geometric loss. SoL augmentation splits a line segment into multiple subparts, which are used to provide auxiliary line data during the training process. Moreover, the matching and geometric loss allow a model to capture additional geometric cues. Compared with TP-LSD-Lite, previously the best real-time LSD method, our model (M-LSD-tiny) achieves competitive performance with 2.5% of model size and an increase of 130.5% in inference speed on GPU. Furthermore, our model runs at 56.8 FPS and 48.6 FPS on the latest Android and iPhone mobile devices, respectively. To the best of our knowledge, this is the first real-time deep LSD available on mobile devices. Our code is available.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑