论文标题
图像掩蔽，用于强大的自我监督单眼估计
Image Masking for Robust Self-Supervised Monocular Depth Estimation
论文作者
论文摘要
自我监督的单眼深度估计是3D场景理解的重要任务。通过单眼的自我运动估计共同学习，已经提出了几种方法来预测准确的像素深度，而无需使用标记的数据。然而，这些方法着重于在没有自然或数字腐败的情况下改善理想条件下的性能。即使是针对特定对象的深度估计，也假定闭塞的总体不存在。这些方法也容易受到对抗性攻击的影响，这是它们在机器人和自主驾驶系统中的可靠部署的相关问题。我们提出了MIMDEPTH，该方法可以适应蒙版的图像建模（MIM）进行自我监督的单眼深度估计。尽管MIM已用于在预训练期间学习可推广的特征，但我们展示了如何将其用于直接训练单眼深度估计。我们的实验表明，对噪声，模糊，天气条件，数字伪像，遮挡以及未靶向和有针对性的对抗性攻击，Mimdepth对噪声，模糊，天气条件，数字伪像，遮挡更为强大。
Self-supervised monocular depth estimation is a salient task for 3D scene understanding. Learned jointly with monocular ego-motion estimation, several methods have been proposed to predict accurate pixel-wise depth without using labeled data. Nevertheless, these methods focus on improving performance under ideal conditions without natural or digital corruptions. The general absence of occlusions is assumed even for object-specific depth estimation. These methods are also vulnerable to adversarial attacks, which is a pertinent concern for their reliable deployment in robots and autonomous driving systems. We propose MIMDepth, a method that adapts masked image modeling (MIM) for self-supervised monocular depth estimation. While MIM has been used to learn generalizable features during pre-training, we show how it could be adapted for direct training of monocular depth estimation. Our experiments show that MIMDepth is more robust to noise, blur, weather conditions, digital artifacts, occlusions, as well as untargeted and targeted adversarial attacks.
