CarLLaVA: Vision language models for camera-only closed-loop driving
This addresses the problem of camera-only autonomous driving for researchers and practitioners, offering a method without complex labels, though it appears incremental as it builds on existing VLM architectures.
The paper tackles autonomous driving by developing CarLLaVA, a vision language model that uses only camera input, achieving state-of-the-art closed-loop performance with a 458% improvement over the previous best and ranking first in the CARLA challenge.
In this technical report, we present CarLLaVA, a Vision Language Model (VLM) for autonomous driving, developed for the CARLA Autonomous Driving Challenge 2.0. CarLLaVA uses the vision encoder of the LLaVA VLM and the LLaMA architecture as backbone, achieving state-of-the-art closed-loop driving performance with only camera input and without the need for complex or expensive labels. Additionally, we show preliminary results on predicting language commentary alongside the driving output. CarLLaVA uses a semi-disentangled output representation of both path predictions and waypoints, getting the advantages of the path for better lateral control and the waypoints for better longitudinal control. We propose an efficient training recipe to train on large driving datasets without wasting compute on easy, trivial data. CarLLaVA ranks 1st place in the sensor track of the CARLA Autonomous Driving Challenge 2.0 outperforming the previous state of the art by 458% and the best concurrent submission by 32.6%.