GEN-SLAM: Generative Modeling for Monocular Simultaneous Localization and Mapping
This work addresses localization and mapping for mobile robots, but it appears incremental as it builds on existing geometric SLAM and deep learning methods.
The paper tackles monocular simultaneous localization and mapping (SLAM) by developing a deep learning system that uses a single camera to estimate topological pose and depth maps for obstacle avoidance, demonstrating effectiveness on simulated and real datasets.
We present a Deep Learning based system for the twin tasks of localization and obstacle avoidance essential to any mobile robot. Our system learns from conventional geometric SLAM, and outputs, using a single camera, the topological pose of the camera in an environment, and the depth map of obstacles around it. We use a CNN to localize in a topological map, and a conditional VAE to output depth for a camera image, conditional on this topological location estimation. We demonstrate the effectiveness of our monocular localization and depth estimation system on simulated and real datasets.