Final project for the Perception in Robotics course

This repository contains the initial stage of the article gradSLAM: Automagically differentiable SLAM implementation.

To run the project you need to install all necessary modules:

pip install -r requirements.txt

To test differentiable Levenberg-Marquardt (LM) solver you can run:

python3 gradLM_exp.py
python3 gradLM_sin.py

or create your own fuction with Function interface (see GradLM.py module).

Also we provide PoseEstimation class for further steps.