DINO Pre-training for Vision-based End-to-end Autonomous Driving

TL;DR

This paper introduces a self-supervised pre-training method using DINO for vision-based autonomous driving, demonstrating improved efficiency over traditional classification pre-training and comparable performance to recent VPR-based methods.

Contribution

The paper proposes using DINO self-distillation for pre-training visual encoders in autonomous driving, moving beyond classification-based methods.

Findings

DINO pre-training outperforms classification-based pre-training in CARLA.

DINO pre-training is as effective as VPRPre in benchmark tests.

Self-supervised pre-training enhances visual understanding for autonomous driving.

Abstract

In this article, we focus on the pre-training of visual autonomous driving agents in the context of imitation learning. Current methods often rely on a classification-based pre-training, which we hypothesise to be holding back from extending capabilities of implicit image understanding. We propose pre-training the visual encoder of a driving agent using the self-distillation with no labels (DINO) method, which relies on a self-supervised learning paradigm.% and is trained on an unrelated task. Our experiments in CARLA environment in accordance with the Leaderboard benchmark reveal that the proposed pre-training is more efficient than classification-based pre-training, and is on par with the recently proposed pre-training based on visual place recognition (VPRPre).