arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2107.06709cs.CVeess.IV
收藏

DVMN:用于深度完成的密集有效性掩码网络

DVMN: Dense Validity Mask Network for Depth Completion

Mannheim University of Applied Sciences(曼海姆应用科学大学)

Laurenz Reichardt, Patrick Mangat, Oliver Wasenmüller

摘要:

LiDAR深度图在各种应用中提供环境指导。然而,这样的深度图通常是稀疏的并且不足以用于诸如自主导航的复杂任务。现有技术的方法使用图像引导的神经网络用于密集深度完井。我们开发了一个引导式卷积神经网络,专注于从稀疏深度图中收集密集和有效的信息。为此,我们引入了一个新的层,该层具有空间变化和内容相关的膨胀,以包括来自稀疏输入的额外数据。此外,我们提出了一个稀疏不变的残留瓶颈块。我们在KITTI深度完成基准上评估了我们的密集有效性掩码网络(DVMN),并获得了最先进的结果。在提交时,我们的网络是使用稀疏不变卷积的领先方法。

英文摘要:

LiDAR depth maps provide environmental guidance in a variety of applications. However, such depth maps are typically sparse and insufficient for complex tasks such as autonomous navigation. State of the art methods use image guided neural networks for dense depth completion. We develop a guided convolutional neural network focusing on gathering dense and valid information from sparse depth maps. To this end, we introduce a novel layer with spatially variant and content-depended dilation to include additional data from sparse input. Furthermore, we propose a sparsity invariant residual bottleneck block. We evaluate our Dense Validity Mask Network (DVMN) on the KITTI depth completion benchmark and achieve state of the art results. At the time of submission, our network is the leading method using sparsity invariant convolution.

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。

↑