AI Paper Atlas

Research map / Multi-agent systems and games / Human-AI and goal-conditioned

Rethinking Closed-Loop Training for Autonomous Driving

Chris Zhang, Runsheng Guo, Wenyuan Zeng, Yuwen Xiong, Binbin Dai, Rui Hu, Mengye Ren, Raquel Urtasun

ECCV 2022 · Multi-agent systems and games › Human-AI and goal-conditioned · approach: Data-driven traffic simulation · 25 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
use trajectory-based output instead of control outputRAINBOW · Targeted scenariospass rate: 0.435 -> 0.814Improved
use trajectory-based output instead of control outputImitation learning · Targeted scenariospass rate: 0.545 -> 0.617Improved
use trajectory-based output instead of control outputPPO · Targeted scenariospass rate: 0.173 -> 0.273Improved
use trajectory-based output instead of control outputA3C · Targeted scenariospass rate: 0.224 -> 0.362Improved
add trajectory value learning with model-based counterfactual loss to RAINBOW+T (TRAVL)RAINBOW+T · Targeted scenariospass rate: 0.814 -> 0.865Improved
add trajectory value learning with model-based counterfactual loss to RAINBOW+T (TRAVL)RAINBOW+T · Targeted scenarioscollision rate: 0.048 -> 0.026Improved
train on targeted instead of free-flow scenarios (ablation)TRAVL · Free-flow test setpass rate: 0.784 -> 0.903Improved
train on targeted instead of free-flow scenarios (ablation)TRAVL · Targeted test setpass rate: 0.696 -> 0.865Improved
train on targeted instead of free-flow scenarios (ablation)RAINBOW+T · Targeted test setpass rate: 0.453 -> 0.815Improved
train on targeted instead of free-flow scenarios (ablation)RAINBOW+T · Free-flow test setpass rate: 0.783 -> 0.885Improved
add behavioral variation to map-only variation in training scenarios (ablation)TRAVL · Targeted scenariospass rate: 0.738 -> 0.872Improved
add map variation to behavioral-only variation in training scenarios (ablation)TRAVL · Targeted scenariospass rate: 0.872 -> 0.865No clear effect

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