arXiv 2108.10869

DROID-SLAM: Deep Visual SLAM for Monocular, Stereo, and RGB-D Cameras

By Zachary Teed and Jia Deng

Published 2021-08-24

Discussion

Read the public discussion and references gathered around this paper.

We introduce DROID-SLAM, a new deep learning based SLAM system. DROID-SLAM consists of recurrent iterative updates of camera pose and pixelwise depth through a Dense Bundle Adjustment layer. DROID-SLAM is accurate, achieving large improvements over prior work, and robust, suffering from substantially fewer catastrophic failures. Despite training on monocular video, it can leverage stereo or RGB-D video to achieve im…

View the original paper on arXiv