/dl2_lfd

Full Robotic Pipeline Codebase for paper: Elaborating on Learned Demonstrations with Temporal Logic Specifications

Primary LanguagePython

Elaborating on Learned Demonstrations with Temporal Logic Specifications

The larger (somewhat messier) supporting code-base for experiments run in the RSS-2020 paper: Elaborating on Learned Demonstrations with Temporal Logic Specifications

Note: For just the specific code relating to differentiable metrics for Linear Temporal Logic, see the submodule LTL-Diff

Setup

This project contains the submodule "LTL_Diff", which constains the language for converting temporal logic statements into differentiable costs. To ensure this submodule is downloaded and updated correctly, make sure you run git clone with the following flag:

git clone --recurse-submodules

Or, if you forget to do this, run the following two commands after cloning the repository:

git submodule init
git submodule update