Depth completion, which aims to generate high-quality dense depth maps from sparse depth maps, has attracted increasing attention in recent years. Previous work usually employs RGB images as guidance, and introduces iterative spatial propagation to refine estimated coarse depth maps. However, most of the propagation refinement methods require several iterations and suffer from a fixed receptive field, which may contain irrelevant and useless information with very sparse input. In this paper, we address these two challenges simultaneously by revisiting the idea of deformable convolution. We propose an effective architecture that leverages deformable kernel convolution as a single-pass refinement module, and empirically demonstrate its superiority. To better understand the function of deformable convolution and exploit it for depth completion, we further systematically investigate a variety of representative strategies. Our study reveals that, different from prior work, deformable convolution needs to be applied on an estimated depth map with a relatively high density for better performance. We evaluate our model on the large-scale KITTI dataset and achieve state-of-the-art level performance in both accuracy and inference speed. Our code is available at https://github.com/AlexSunNik/ReDC.

翻译：深度补全旨在从稀疏深度图中生成高质量稠密深度图，近年来受到越来越多的关注。先前的工作通常以RGB图像为指导，并引入迭代式空间传播以细化估计的粗略深度图。然而，大多数传播细化方法需要多次迭代，且受限于固定的感受野，在输入极为稀疏时可能包含无关和无用的信息。本文通过重新审视可变形卷积的思路，同时解决这两个挑战。我们提出一种高效架构，将可变形核卷积作为单次通过（single-pass）细化模块，并通过实验证明其优越性。为更深入理解可变形卷积的功能并利用其进行深度补全，我们进一步系统研究了多种代表性策略。研究表明，与先前工作不同，可变形卷积需要应用于密度相对较高的估计深度图上才能取得更优性能。我们在大规模KITTI数据集上评估了模型，在精度和推理速度上均达到了当前最优水平。代码已开源在https://github.com/AlexSunNik/ReDC。