MapDiffusion: Generative Diffusion for Vectorized Online HD Map Construction and Uncertainty Estimation in Autonomous Driving
Thomas Monninger, Zihan Zhang, Zhipeng Mo, Md Zafar Anwar, Steffen Staab, Sihao Ding

TL;DR
MapDiffusion introduces a generative diffusion model for vectorized HD map construction in autonomous driving, capturing uncertainty and improving robustness over deterministic methods.
Contribution
It presents a novel diffusion-based approach that models the full distribution of possible maps, enabling uncertainty estimation and improved accuracy in online map construction.
Findings
Achieves 5% better performance than baseline in single-sample tests.
Sample aggregation improves overall map prediction accuracy.
Uncertainty estimates are higher in occluded or ambiguous regions.
Abstract
Autonomous driving requires an understanding of the static environment from sensor data. Learned Bird's-Eye View (BEV) encoders are commonly used to fuse multiple inputs, and a vector decoder predicts a vectorized map representation from the latent BEV grid. However, traditional map construction models provide deterministic point estimates, failing to capture uncertainty and the inherent ambiguities of real-world environments, such as occlusions and missing lane markings. We propose MapDiffusion, a novel generative approach that leverages the diffusion paradigm to learn the full distribution of possible vectorized maps. Instead of predicting a single deterministic output from learned queries, MapDiffusion iteratively refines randomly initialized queries, conditioned on a BEV latent grid, to generate multiple plausible map samples. This allows aggregating samples to improve prediction…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
