DistillDrive: End-to-End Multi-Mode Autonomous Driving Distillation by Isomorphic Hetero-Source Planning Model

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Yu, Rui, Zhang, Xianghang, Zhao, Runkai, Yan, Huaicheng, Wang, Meng
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915433784803328
author Yu, Rui
Zhang, Xianghang
Zhao, Runkai
Yan, Huaicheng
Wang, Meng
author_facet Yu, Rui
Zhang, Xianghang
Zhao, Runkai
Yan, Huaicheng
Wang, Meng
contents End-to-end autonomous driving has been recently seen rapid development, exerting a profound influence on both industry and academia. However, the existing work places excessive focus on ego-vehicle status as their sole learning objectives and lacks of planning-oriented understanding, which limits the robustness of the overall decision-making prcocess. In this work, we introduce DistillDrive, an end-to-end knowledge distillation-based autonomous driving model that leverages diversified instance imitation to enhance multi-mode motion feature learning. Specifically, we employ a planning model based on structured scene representations as the teacher model, leveraging its diversified planning instances as multi-objective learning targets for the end-to-end model. Moreover, we incorporate reinforcement learning to enhance the optimization of state-to-decision mappings, while utilizing generative modeling to construct planning-oriented instances, fostering intricate interactions within the latent space. We validate our model on the nuScenes and NAVSIM datasets, achieving a 50\% reduction in collision rate and a 3-point improvement in closed-loop performance compared to the baseline model. Code and model are publicly available at https://github.com/YuruiAI/DistillDrive
format Preprint
id arxiv_https___arxiv_org_abs_2508_05402
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle DistillDrive: End-to-End Multi-Mode Autonomous Driving Distillation by Isomorphic Hetero-Source Planning Model
Yu, Rui
Zhang, Xianghang
Zhao, Runkai
Yan, Huaicheng
Wang, Meng
Robotics
Computer Vision and Pattern Recognition
End-to-end autonomous driving has been recently seen rapid development, exerting a profound influence on both industry and academia. However, the existing work places excessive focus on ego-vehicle status as their sole learning objectives and lacks of planning-oriented understanding, which limits the robustness of the overall decision-making prcocess. In this work, we introduce DistillDrive, an end-to-end knowledge distillation-based autonomous driving model that leverages diversified instance imitation to enhance multi-mode motion feature learning. Specifically, we employ a planning model based on structured scene representations as the teacher model, leveraging its diversified planning instances as multi-objective learning targets for the end-to-end model. Moreover, we incorporate reinforcement learning to enhance the optimization of state-to-decision mappings, while utilizing generative modeling to construct planning-oriented instances, fostering intricate interactions within the latent space. We validate our model on the nuScenes and NAVSIM datasets, achieving a 50\% reduction in collision rate and a 3-point improvement in closed-loop performance compared to the baseline model. Code and model are publicly available at https://github.com/YuruiAI/DistillDrive
title DistillDrive: End-to-End Multi-Mode Autonomous Driving Distillation by Isomorphic Hetero-Source Planning Model
topic Robotics
Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2508.05402