Assignment 1 - Benchmarking public VINS/LIOM methods and datasets

Instructions:

Pick one open-sourced Visual-Inertial Localization method, and one open-sourced Lidar-Inertial Localization method in the literature.

Test the methods with all 18 sequences in NTU VIRAL datasets. Record the estimated trajectory and calculate the Absolute Trajectory Error (ATE).

Note: You can write a linux script to automate all 36 experiments. 

Accessment:
1. Summarize all results in a pdf. Add some remarks on the accuracy and robustness and your explanation. Add one or two examples of when loop closure reduces localization drift.

2. Demonstrate the experiments in person on the review day.

Peer-review:

Please meet (over Zoom is OK but IRL best) and discuss as part of the peer-review. It is OK to adjust your solution based on what you learned here.