autonomousvision/king

generate gif

Closed this issue · 2 comments

Thank you for releasing this project.
But how should I generate the gif in your home page from the generation_results/scenario_records.json.
Thank you very much.

Hi! Generating a .gif exactly as the teaser on the landing page of this repository is a bit involved. However, you can use python tools/visualize_scenarios.py, which is a lightweight script that draws .gifs of the generated scenarios in a simple BEV representation. This is also what we used during development for this project.

To use it, you need to spin up a CARLA server in the background. Make sure the argument --scenario_log_dir points to the directory where your generation results are located. Additionally you may need to set the environment variables for CARLA's python API as in run_generation.sh (l.3-7).

Okay I got it. Thank you for your reply!