Skip to content

Latest commit

 

History

History

socnavbench

Folders and files

NameName
Last commit message
Last commit date

parent directory

..
README.md
README.md
 
 
brne.py
brne.py
 
 
joystick_brne.py
joystick_brne.py
 
 
joystick_client.py
joystick_client.py
 
 
joystick_cvm.py
joystick_cvm.py
 
 

README.md

BRNE implementation for SocNavBench

Author: Max Muchen Sun

Social Navigation Benchmark utility (SocNavBench) is a codebase for benchmarking robot planning algorithms against various episodes of containing multi-agent environments, developed at Carneige Mellon University. More information regarding the benchmark framework can be found here.

Below are the implementations of BRNE for the SocNavBench environment and how to install them. The implementations include the BRNE algorihtm (brne) and the constant velocity baseline (cvm).

To install:

  1. Install SocNavBench following the instructions.
  2. Copy the file joystick_client.py into the directory (root dir of SovNavBench)/joystick.
  3. Copy the files brne.py, joystick_brne.py, and joystick_cvm.py into the directory (root dir of SovNavBench)/joystick/joystick_py.

To use:

  1. Follow the usage instructions of SovNavBench.
  2. Use the argument --algo "brne" or --algo "cvm" to run BRNE or the constant velocity baseline.