AI Paper Atlas

Research map / Label-efficient and robust learning / Recognition and long-tail

DriveAdapter: Breaking the Coupling Barrier of Perception and Planning in End-to-End Autonomous Driving

Xiaosong Jia, Yulu Gao, Li Chen, Junchi Yan, Patrick Langechuan Liu, Hongyang Li

ICCV 2023 · Label-efficient and robust learning › Recognition and long-tail · approach: Dense features and foundation priors · 107 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add adapters with feature alignment between student perception and frozen teacherstudent + frozen Roach · CARLA Town05 LongDriving Score: 8.9 -> 61.7Improved
add feature alignment loss (ablation)DriveAdapter · CARLA Town05 LongDriving Score: 45.4 -> 61.7Improved
add mask on feature alignment for rule-overridden cases (ablation)DriveAdapter · CARLA Town05 LongDriving Score: 56.9 -> 61.7Improved
add action loss (ablation)DriveAdapter · CARLA Town05 LongDriving Score: 47.1 -> 61.7Improved
place adapters at all stages rather than only early stage (ablation)DriveAdapter · CARLA Town05 LongDriving Score: 47.2 -> 61.7Improved
place adapters at all stages rather than only late stage (ablation)DriveAdapter · CARLA Town05 LongDriving Score: 54.3 -> 61.7Improved
feed raw BEV feature to adapters (ablation)DriveAdapter · CARLA Town05 LongDriving Score: 34.8 -> 61.7Improved
freeze the teacher model instead of unfreezing (ablation)DriveAdapter · CARLA Town05 LongDriving Score: 49.0 -> 61.7Improved
add TCP-style trajectory prediction headDriveAdapter · CARLA Town05 LongDriving Score: 61.7 -> 65.9Improved
train with 2M frames over 8 towns instead of 189K framesDriveAdapter + TCP · CARLA Longest6Driving Score: 62.0 -> 71.4Improved
make student regress deeper teacher features instead of BEV segmentation (ablation)student model + frozen teacher · CARLA Town05 LongDriving Score: 8.9 -> 39.0 (Latent)Improved

Later papers that compared with this method

6 later accepted papers report DriveAdapter as a baseline in their results tables; 6 of them report a better number on most of the results they share with it.

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy