Bench2Drive-Robust is a new closed-loop benchmark that evaluates end-to-end autonomous driving models under deployment perturbations from camera failures, ego-state errors, and compute delays, showing substantial performance degradation beyond image-level tests.
hub Baseline reference
NuPlan: A closed-loop ML-based planning benchmark for autonomous vehicles
Baseline reference. 75% of citing Pith papers use this work as a benchmark or comparison.
34
Pith papers citing it
Baseline
75% of classified citations
abstract
In this work, we propose the world's first closed-loop ML-based planning benchmark for autonomous driving. While there is a growing body of ML-based motion planners, the lack of established datasets and metrics has limited the progress in this area. Existing benchmarks for autonomous vehicle motion prediction have focused on short-term motion forecasting, rather than long-term planning. This has led previous works to use open-loop evaluation with L2-based metrics, which are not suitable for fairly evaluating long-term planning. Our benchmark overcomes these limitations by introducing a large-scale driving dataset, lightweight closed-loop simulator, and motion-planning-specific metrics. We provide a high-quality dataset with 1500h of human driving data from 4 cities across the US and Asia with widely varying traffic patterns (Boston, Pittsburgh, Las Vegas and Singapore). We will provide a closed-loop simulation framework with reactive agents and provide a large set of both general and scenario-specific planning metrics. We plan to release the dataset at NeurIPS 2021 and organize benchmark challenges starting in early 2022.
hub tools
citation-role summary
dataset 8
background 3
baseline 1
citation-polarity summary
representative citing papers
MDrive benchmark shows multi-agent cooperative driving systems generally outperform single-agent ones in closed-loop settings but perception sharing does not always improve planning and negotiation can harm performance in complex traffic.
CROWD is a new global dataset of 51,753 continuous urban dashcam segments spanning over 20,000 hours from 238 countries, with manual labels and automated object detections for routine driving analysis.
C-TRAIL combines LLM commonsense with a dual-trust mechanism and Dirichlet-weighted Monte Carlo Tree Search to improve trajectory planning accuracy and safety in autonomous driving.
KITScenes LongTail supplies multimodal driving data and multilingual expert reasoning traces to benchmark models on rare scenarios beyond basic safety metrics.
ReCogDrive unifies VLM scene understanding with a diffusion planner reinforced by DiffGRPO to reach state-of-the-art results on NAVSIM and Bench2Drive benchmarks.
BeyondDrive augments imitation learning with synthesized safety-critical negative trajectories and a repulsive loss to improve safety in autonomous driving, reporting 89.7 PDMS on NAVSIMv1 and generalization to other models.
CoWorld-VLA extracts semantic, geometric, dynamic, and trajectory expert tokens from multi-source supervision and feeds them into a diffusion-based hierarchical planner, achieving competitive collision avoidance and trajectory accuracy on the NAVSIM v1 benchmark.
Smaller end-to-end autonomous driving models achieve optimal 3-second trajectory prediction accuracy at lower or intermediate temporal sampling frequencies, whereas larger VLA-style models perform best at the highest frequencies across Waymo, nuScenes, and PAVE datasets.
DriveFuture achieves SOTA results on NAVSIM by conditioning latent world model states on future predictions to directly inform trajectory planning.
SceneFactory delivers a batched GPU platform for physics-based multi-agent autonomous driving simulation that achieves 127x higher throughput than non-vectorized PhysX while supporting articulated dynamics and road-condition friction.
Response times modeled as drift-diffusion processes enable consistent estimation of population-average preferences from heterogeneous anonymous binary choices.
ReflectDrive-2 combines masked discrete diffusion with RL-aligned self-editing to generate and refine driving trajectories, reaching 91.0 PDMS on NAVSIM camera-only and 94.8 in best-of-6.
ProDrive couples a query-centric planner with a BEV world model for end-to-end ego-environment co-evolution, enabling future-outcome assessment that improves safety and efficiency over reactive baselines on NAVSIM v1.
OneDrive unifies heterogeneous decoding in a single VLM transformer decoder for end-to-end driving, achieving 0.28 L2 error and 0.18 collision rate on nuScenes plus 86.8 PDMS on NAVSIM.
Mosaic integrates rule-based and learned planners via arbitration graphs to set new state-of-the-art scores on nuPlan and interPlan benchmarks while cutting at-fault collisions by 30%.
The primary OL-CL gap in end-to-end autonomous driving arises from objective mismatch creating structural inability to model reactive behaviors, which a test-time adaptation method can mitigate.
E² uses transport-regularized sparse control on learned reverse-time SDEs with topology-driven selection and Topological Anchoring to generate realistic adversarial scenarios, improving collision discovery by 9.01% on nuScenes and up to 21.43% on nuPlan while enabling closed-loop robustness gains.
The paper introduces Hyper Diffusion Planner (HDP), a diffusion-based E2E AD framework that identifies insights on loss space, trajectory representation and data scaling, adds RL post-training, and reports 10x performance gains over 200 km of real-world testing across 6 scenarios.
DriveLaW unifies video world modeling and trajectory planning by injecting video-generator latents into a diffusion planner, achieving SOTA video prediction and a new record on the NAVSIM planning benchmark.
OMEGA guides diffusion sampling with per-step constrained optimization and game-theoretic adversarial modeling to generate physically valid and interactive driving scenes, raising valid scene ratios from 32% to 72% and producing 5x more near-collisions.
EnDfuser replaces point-estimate trajectory planning with ensemble diffusion in a single attention-pooling transformer module to model posterior trajectory uncertainty and improve safety in end-to-end autonomous driving.
LiloDriver uses LLMs and memory-augmented planning in a four-stage pipeline to outperform rule-based and learning-based methods on both common and rare scenarios in the nuPlan benchmark.
LAW introduces a self-supervised prediction task on latent scene features that boosts end-to-end driving performance on nuScenes, NAVSIM, and CARLA benchmarks.
citing papers explorer
-
Bench2Drive-Robust: Benchmarking Closed-Loop Autonomous Driving under Deployment Perturbations
Bench2Drive-Robust is a new closed-loop benchmark that evaluates end-to-end autonomous driving models under deployment perturbations from camera failures, ego-state errors, and compute delays, showing substantial performance degradation beyond image-level tests.
-
MDrive: Benchmarking Closed-Loop Cooperative Driving for End-to-End Multi-agent Systems
MDrive benchmark shows multi-agent cooperative driving systems generally outperform single-agent ones in closed-loop settings but perception sharing does not always improve planning and negotiation can harm performance in complex traffic.
-
A global dataset of continuous urban dashcam driving
CROWD is a new global dataset of 51,753 continuous urban dashcam segments spanning over 20,000 hours from 238 countries, with manual labels and automated object detections for routine driving analysis.
-
C-TRAIL: A Commonsense World Framework for Trajectory Planning in Autonomous Driving
C-TRAIL combines LLM commonsense with a dual-trust mechanism and Dirichlet-weighted Monte Carlo Tree Search to improve trajectory planning accuracy and safety in autonomous driving.
-
LongTail Driving Scenarios with Reasoning Traces: The KITScenes LongTail Dataset
KITScenes LongTail supplies multimodal driving data and multilingual expert reasoning traces to benchmark models on rare scenarios beyond basic safety metrics.
-
ReCogDrive: A Reinforced Cognitive Framework for End-to-End Autonomous Driving
ReCogDrive unifies VLM scene understanding with a diffusion planner reinforced by DiffGRPO to reach state-of-the-art results on NAVSIM and Bench2Drive benchmarks.
-
Beyond Imitation: Learning Safe End-to-End Autonomous Driving from Hard Negatives
BeyondDrive augments imitation learning with synthesized safety-critical negative trajectories and a repulsive loss to improve safety in autonomous driving, reporting 89.7 PDMS on NAVSIMv1 and generalization to other models.
-
CoWorld-VLA: Thinking in a Multi-Expert World Model for Autonomous Driving
CoWorld-VLA extracts semantic, geometric, dynamic, and trajectory expert tokens from multi-source supervision and feeds them into a diffusion-based hierarchical planner, achieving competitive collision avoidance and trajectory accuracy on the NAVSIM v1 benchmark.
-
Temporal Sampling Frequency Matters: A Capacity-Aware Study of End-to-End Driving Trajectory Prediction
Smaller end-to-end autonomous driving models achieve optimal 3-second trajectory prediction accuracy at lower or intermediate temporal sampling frequencies, whereas larger VLA-style models perform best at the highest frequencies across Waymo, nuScenes, and PAVE datasets.
-
DriveFuture: Future-Aware Latent World Models for Autonomous Driving
DriveFuture achieves SOTA results on NAVSIM by conditioning latent world model states on future predictions to directly inform trajectory planning.
-
SceneFactory: GPU-Accelerated Multi-Agent Driving Simulation with Physics-Based Vehicle Dynamics
SceneFactory delivers a batched GPU platform for physics-based multi-agent autonomous driving simulation that achieves 127x higher throughput than non-vectorized PhysX while supporting articulated dynamics and road-condition friction.
-
Response Time Enhances Alignment with Heterogeneous Preferences
Response times modeled as drift-diffusion processes enable consistent estimation of population-average preferences from heterogeneous anonymous binary choices.
-
ReflectDrive-2: Reinforcement-Learning-Aligned Self-Editing for Discrete Diffusion Driving
ReflectDrive-2 combines masked discrete diffusion with RL-aligned self-editing to generate and refine driving trajectories, reaching 91.0 PDMS on NAVSIM camera-only and 94.8 in best-of-6.
-
ProDrive: Proactive Planning for Autonomous Driving via Ego-Environment Co-Evolution
ProDrive couples a query-centric planner with a BEV world model for end-to-end ego-environment co-evolution, enabling future-outcome assessment that improves safety and efficiency over reactive baselines on NAVSIM v1.
-
OneDrive: Unified Multi-Paradigm Driving with Vision-Language-Action Models
OneDrive unifies heterogeneous decoding in a single VLM transformer decoder for end-to-end driving, achieving 0.28 L2 error and 0.18 collision rate on nuScenes plus 86.8 PDMS on NAVSIM.
-
Mosaic: An Extensible Framework for Composing Rule-Based and Learned Motion Planners
Mosaic integrates rule-based and learned planners via arbitration graphs to set new state-of-the-art scores on nuPlan and interPlan benchmarks while cutting at-fault collisions by 30%.
-
BridgeSim: Unveiling the OL-CL Gap in End-to-End Autonomous Driving
The primary OL-CL gap in end-to-end autonomous driving arises from objective mismatch creating structural inability to model reactive behaviors, which a test-time adaptation method can mitigate.
-
Evaluation as Evolution: Transforming Adversarial Diffusion into Closed-Loop Curricula for Autonomous Vehicles
E² uses transport-regularized sparse control on learned reverse-time SDEs with topology-driven selection and Topological Anchoring to generate realistic adversarial scenarios, improving collision discovery by 9.01% on nuScenes and up to 21.43% on nuPlan while enabling closed-loop robustness gains.
-
Unleashing the Potential of Diffusion Models for End-to-End Autonomous Driving
The paper introduces Hyper Diffusion Planner (HDP), a diffusion-based E2E AD framework that identifies insights on loss space, trajectory representation and data scaling, adds RL post-training, and reports 10x performance gains over 200 km of real-world testing across 6 scenarios.
-
DriveLaW:Unifying Planning and Video Generation in a Latent Driving World
DriveLaW unifies video world modeling and trajectory planning by injecting video-generator latents into a diffusion planner, achieving SOTA video prediction and a new record on the NAVSIM planning benchmark.
-
Optimization-Guided Diffusion for Interactive Scene Generation
OMEGA guides diffusion sampling with per-step constrained optimization and game-theoretic adversarial modeling to generate physically valid and interactive driving scenes, raising valid scene ratios from 32% to 72% and producing 5x more near-collisions.
-
Using Ensemble Diffusion to Estimate Uncertainty for End-to-End Autonomous Driving
EnDfuser replaces point-estimate trajectory planning with ensemble diffusion in a single attention-pooling transformer module to model posterior trajectory uncertainty and improve safety in end-to-end autonomous driving.
-
LiloDriver: A Lifelong Learning Framework for Closed-loop Motion Planning in Long-tail Autonomous Driving Scenarios
LiloDriver uses LLMs and memory-augmented planning in a four-stage pipeline to outperform rule-based and learning-based methods on both common and rare scenarios in the nuPlan benchmark.
-
Enhancing End-to-End Autonomous Driving with Latent World Model
LAW introduces a self-supervised prediction task on latent scene features that boosts end-to-end driving performance on nuScenes, NAVSIM, and CARLA benchmarks.
-
Hydra-MDP: End-to-end Multimodal Planning with Multi-target Hydra-Distillation
Hydra-MDP uses multi-teacher distillation and a multi-head decoder to learn diverse, metric-specific trajectories in an end-to-end autonomous-driving planner, winning the Navsim challenge.
-
BitTP: The Lightweight Trajectory Prediction Model with BitLLM for Edge-Devices
BitTP applies weight-only 1.58-bit quantization to LLM trajectory predictors, claiming improved ADE/FDE over BF16 baseline with reduced resource demands on edge devices.
-
Distill to Think, Foresee to Act: Cognitive-Physical Reinforcement Learning for Autonomous Driving
CoPhy is a new RL framework that distills VLM cognition into BEV encoders, adds an auto-regressive BEV world model for action-conditioned future prediction, and optimizes policies via GRPO with dual physical-cognitive rewards, claiming SOTA on NAVSIM v1/v2.
-
HEAT: Heterogeneous End-to-End Autonomous Driving via Trajectory-Guided World Models
HEAT uses a trajectory-driven learning paradigm and a world model predicting future latent features from ego actions to enable a single unified end-to-end autonomous driving model to perform well across heterogeneous domains on nuScenes, NAVSIM, and Waymo benchmarks.
-
RLFTSim: Realistic and Controllable Multi-Agent Traffic Simulation via Reinforcement Learning Fine-Tuning
RLFTSim uses RL fine-tuning on a pre-trained model with a balanced reward to align traffic simulator rollouts to real data distributions and distill goal-conditioned controllability, reporting SOTA realism on the Waymo Open Motion Dataset.
-
DriveSafer: End-to-End Autonomous Driving with Safety Guidance
DriveSafer reduces catastrophic failures (PDMS=0) by 48% and drivable-area compliance failures by over 65% versus DiffusionDrive on the NAVSIM benchmark by combining training-time safety constraints with inference-time guidance.
-
Causality-Aware End-to-End Autonomous Driving via Ego-Centric Joint Scene Modeling
CaAD adds ego-centric joint-causal modeling and causality-aware policy alignment to end-to-end driving, reporting Driving Score 87.53 and PDMS 91.1 on Bench2Drive and NAVSIM.
-
Artificial Intelligence for Modeling and Simulation of Mixed Automated and Human Traffic
This survey synthesizes AI techniques for mixed autonomy traffic simulation and introduces a taxonomy spanning agent-level behavior models, environment-level methods, and cognitive/physics-informed approaches.
-
CHARMS: A Cognitive Hierarchical Agent for Reasoning and Motion Stylization in Autonomous Driving
CHARMS applies Level-k game theory and Poisson cognitive hierarchy theory to autonomous driving agents via a two-stage RL-then-SFT pipeline for human-like decisions and realistic scenario generation.
- DIVER: Reinforced Diffusion Breaks Imitation Bottlenecks in End-to-End Autonomous Driving