Toy implementation of the Visual Simultaneous Localization and Mapping (SLAM) algorithm in Rust with a trick up its sleeve. This implementation of VSLAM utilizes a dense depth model proposed in the paper High Quality Monocular Depth Estimation via Transfer Learning (arXiv:1812.11941) for the mapping component.
Make sure you have cargo and rust installed.
To run this make sure you have Xcode 14 installed. To build this project run:
$ cargo buildTo run this project:
$ cargo run datasets/vids/test_ohio.mp4