VINS-Fusion

Supports various configurations like mono camera+IMU, stereo cameras+IMU, and stereo-only, as demonstrated with EuRoC and KITTI datasets, allowing adaptation to different hardware setups.

Automatically calibrates spatial and temporal offsets between cameras and IMU during operation, reducing manual pre-calibration efforts, as highlighted in the features list.

Ranks as the top open-source stereo algorithm on the KITTI Odometry Benchmark, ensuring high precision for autonomous vehicle applications.

Incorporates visual loop detection to reduce drift over long trajectories, improving long-term odometry accuracy, as shown in demo videos with green and red paths.