MomWorld: Momentum-Aware Latent World Model for Long-Horizon Autonomous Driving
cs.RO
Submitted: 2026-09-27
Updated: 2026-09-30
Code: https://github.com/modaxiansheng/MomWorld
License: http://creativecommons.org/licenses/by/4.0/
The gist: Long-horizon planning enables autonomous vehicles to anticipate scene evolution and potential risks, supporting safe and stable decisions in complex interactions.
Terminology
Abstract
Long-horizon planning enables autonomous vehicles to anticipate scene evolution and potential risks, supporting safe and stable decisions in complex interactions. However, existing methods struggle to propagate motion trends from observed history into the future. Long rollouts based on a single latent state may further attenuate useful dynamics, retain stale motion patterns, and disrupt reliable near-term plans. We introduce MomWorld, a momentum-aware latent world model for long-horizon planning. MomWorld extracts scene motion trends from historical-to-current observations and propagates latent momentum into future horizons, jointly predicting future configuration and momentum states. A learnable momentum persistence mechanism preserves stable trends, scene-conditioned momentum updates adapt future dynamics, and a scene-adaptive reset gate suppresses stale momentum under abrupt changes. We further propose MoFlow, a momentum-conditioned flow-matching module that refines a base trajectory to align with the predicted future scene evolution in only a few integration steps, with a horizon-aware residual fusion that preserves near-term planning stability while permitting stronger long-range corrections. Extensive experiments on NAVSIM, nuScenes and Bench2Drive demonstrate that MomWorld improves long-horizon planning consistency and reduces the average collision rate by 12.2% relative to MomAD over a 6-second planning horizon.
Sources
- VADv2: End-to-End Vectorized Autonomous Driving via Probabilistic Planning
- PV-WM: A Heterogeneous Micro-Macro World Model for Articulated Pedestrian-Vehicle Co-Rollout
- Driver-WM: A Driver-Centric Traffic-Conditioned Latent World Model for In-Cabin Dynamics Rollout
- DriveFine: Refining-Augmented Masked Diffusion VLA for Precise and Robust Driving
- DriveFuture: Future-Aware Latent World Models for Autonomous Driving
- Hydra-MDP++: Advancing End-to-End Driving via Expert-Guided Hydra-Distillation
- End-to-End Driving with Online Trajectory Evaluation via BEV World Model
- ReCogDrive: A Reinforced Cognitive Framework for End-to-End Autonomous Driving
- Hydra-MDP: End-to-end Multimodal Planning with Multi-target Hydra-Distillation
- Generalized Trajectory Scoring for End-to-end Multimodal Planning
- Fully Unified Motion Planning for End-to-End Autonomous Driving
- DriveWorld-VLA: Unified Latent-Space World Modeling with Vision-Language-Action for Autonomous Driving
- MindDrive: An All-in-One Framework Bridging World Models and Vision-Language Model for End-to-End Autonomous Driving
- SparseDriveV2: Scoring is All You Need for End-to-End Autonomous Driving
- Latent-WAM: Latent World Action Modeling for End-to-End Autonomous Driving
- Challenger: Affordable Adversarial Driving Video Generation
- DriveSuprim: Towards Precise Trajectory Selection for End-to-End Planning
- DRAMA: An Efficient End-to-end Motion Planner for Autonomous Driving with Mamba
- AutoVLA: A Vision-Language-Action Model for End-to-End Autonomous Driving with Adaptive Reasoning and Reinforcement Fine-Tuning
- SpanVLA: Learning from Negative-Recovery Samples with Fast Action Bridging for Vision-Language-Action Model
Related papers
- FMT x: An Efficient and Asymptotically Optimal Extension of the Fast Marching Tree for Dynamic Replanning
- MPCFormer: A physics-informed data-driven approach for explainable socially-aware autonomous driving
- RoboLab: A High-Fidelity Simulation Benchmark for Analysis of Task Generalist Policies
- ARC: A Reasoning Recipe for Robot Foundation Models
- HRDexDB: A 4D Dexterous Grasping Dataset Across Human and Multiple Robot Embodiments
- APT: Action Expert Pretraining Improves Instruction Generalization of Vision-Language-Action Policies