Finding a Baseline for Camera-Only Autonomous Driving in CARLA

[Akumar201]

Hello, I’m trying to find a baseline for autonomous driving in CARLA that I can use with my image compression algorithm. Could you please suggest something? I’m new to CARLA, so I’d appreciate any existing policies or methods that use only camera input as a baseline.

[Kait0]

LatentTransFuser or TCP are common efficient image only baselines.

SimLingo is a recent SOTA model witch only uses camera inputs.