arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1702.02706cs.CV
收藏

用于单目深度图预测的半监督深度学习

Semi-Supervised Deep Learning for Monocular Depth Map Prediction

RWTH Aachen University(亚琛工业大学)

Yevhen Kuznietsov, Jörg Stückler, Bastian Leibe

摘要:

受监督的深度学习经常受到缺乏足够的训练数据的困扰。具体地,在单目深度图预测的上下文中,几乎不可能在现实的动态户外环境中确定密集的地面实况深度图像。例如,当使用LiDAR传感器时,距离测量中存在噪声,传感器之间的校准不可能是完美的,并且测量通常比相机图像稀疏得多。在本文中,我们提出了一种新的方法来从单目图像中预测深度图,以半监督的方式学习。虽然我们使用稀疏地面实况深度进行监督学习,但我们也使用直接图像对齐损失来强制我们的深度网络在立体设置中产生照片一致性密集深度图。在实验中,我们展示了优越的性能,从单个图像的深度图预测相比,国家的最先进的方法。

英文摘要:

Supervised deep learning often suffers from the lack of sufficient training data. Specifically in the context of monocular depth map prediction, it is barely possible to determine dense ground truth depth images in realistic dynamic outdoor environments. When using LiDAR sensors, for instance, noise is present in the distance measurements, the calibration between sensors cannot be perfect, and the measurements are typically much sparser than the camera images. In this paper, we propose a novel approach to depth map prediction from monocular images that learns in a semi-supervised way. While we use sparse ground-truth depth for supervised learning, we also enforce our deep network to produce photoconsistent dense depth maps in a stereo setup using a direct image alignment loss. In experiments we demonstrate superior performance in depth map prediction from single images compared to the state-of-the-art methods.

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。

↑