Self-supervised monocular depth estimation with self-distillation and dense skip connection

Xuezhi Xiang, Wei Li, Yao Wang, Abdulmotaleb El-Saddik

Abstract: Highlights•We propose a successive depth map self-distillation loss for self-supervised monocular depth estimation.•We propose a dense skip connection strategy to improve the depth estimation effect of the depth network.•We validate the proposed method’s effectiveness on the KITTI dataset and NYUv2 dataset, achieving state-of-the-art performance.