Planning with Adaptive World Models for Autonomous Driving

arXiv 24.6 2024 67.6 method, application

TLDR

AdaptiveDriver uses BehaviorNet to model agent behaviors and MPC planning, achieving SOTA on nuPlan benchmark.

Reasoning

The paper introduces a novel approach to motion planning by learning adaptive world models that capture unique driving behaviors per city, validated on the real-world nuPlan benchmark. Strengths include addressing multi-agent interactions and achieving state-of-the-art results; weaknesses include limited evaluation to a single benchmark and lack of discussion on generalization to other environments.

Read-first score

Read-first score 67.6, weighted from topical fit, citation, graph, method, reproducibility, and recency signals. Original total remains 27.

Methodology quality 25%

100

Screens visible abstract and analysis fields for experiment, dataset, baseline, metric, and limitation evidence. markers=analysis,benchmark,dataset,evaluation,experiment,result

Reproducibility 25%

Screens links and visible text for paper, code, dataset, artifact, and repository signals. pdf=True; code=True; dataset=False; markers=dataset,github

Recency 8%

75.1

Uses a gentle age decay so recent papers surface without erasing older foundations. 2024

Topical relevance 42%

38.6

Uses existing LLM keyword relevance scores normalized to 0-100. world model,world simulator,generative world model,interactive world model,video world model,world dynamics prediction,model-based reinforcement learning world model

Field roles

Methodology anchorReproducibility anchor

Rank sensitivity

Stability: volatile; rank range: 536.

Keyword Scores

world model

world dynamics prediction

interactive world model

world simulator

model-based reinforcement learning world model

generative world model

video world model

Deep Analysis

Innovations

BehaviorNet: a graph convolutional neural network (GCNN) that predicts reactive agent behaviors by predicting parameters of an agent's motion controller rather than directly predicting its spacetime trajectory
AdaptiveDriver: a model-predictive control (MPC) based planner that unrolls different world models conditioned on BehaviorNet's predictions
Analysis of nuPlan's recorded logs revealing that each city has its own unique driving behaviors, motivating the need for adaptive planners

Methodology

The paper uses the nuPlan benchmark, which augments real-world driving logs with closed-loop simulation logic. They learn BehaviorNet, a GCNN that predicts parameters of an agent's motion controller from recently-observed agent histories. Then AdaptiveDriver uses MPC to unroll different world models conditioned on BehaviorNet's predictions. Evaluation is performed on the nuPlan closed-loop planning benchmark.

Key Results

AdaptiveDriver achieves state-of-the-art results on the nuPlan closed-loop planning benchmark, improving over prior work by 2% on Test-14 Hard R-CLS, and generalizes even when evaluated on never-before-seen cities.

Resources

Paper Code