Multi-Session SLAM with Differentiable Wide-Baseline Pose Optimization

arXiv:2404.15263v1 Announce Type: new
Abstract: We introduce a new system for Multi-Session SLAM, which tracks camera motion across multiple disjoint videos under a single global reference. Our approach couples the prediction of optical flow with solver layers to estimate camera pose. The backbone is trained end-to-end using a novel differentiable solver for wide-baseline two-view pose. The full system can connect disjoint sequences, perform visual odometry, and global optimization. Compared to existing approaches, our design is accurate and robust to catastrophic …

abstract arxiv cs.cv differentiable flow global multiple novel optical optical flow optimization prediction reference session slam solver type videos