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 failures. Code is available at github.com/princeton-vl/MultiSlam_DiffPose
@article{arxiv.2404.15263,
title = {Multi-Session SLAM with Differentiable Wide-Baseline Pose Optimization},
author = {Lahav Lipson and Jia Deng},
journal= {arXiv preprint arXiv:2404.15263},
year = {2024}
}