NRSeg: Noise-Resilient Learning for BEV Semantic Segmentation via Driving World Models

TL;DR

This paper introduces NRSeg, a noise-resilient learning framework for BEV semantic segmentation in autonomous driving, leveraging synthetic data and novel metrics to improve robustness and accuracy in semi-supervised and unsupervised settings.

Contribution

NRSeg proposes a new framework with a perspective-geometry consistency metric, a dual-distribution prediction approach, and a hierarchical local semantic exclusion module for robust BEV segmentation.

Findings

Achieves state-of-the-art mIoU improvements of 13.8% and 11.4% in unsupervised and semi-supervised tasks.

Effectively utilizes synthetic data despite generation noise.

Demonstrates robustness and superior performance over existing methods.

Abstract

Birds' Eye View (BEV) semantic segmentation is an indispensable perception task in end-to-end autonomous driving systems. Unsupervised and semi-supervised learning for BEV tasks, as pivotal for real-world applications, underperform due to the homogeneous distribution of the labeled data. In this work, we explore the potential of synthetic data from driving world models to enhance the diversity of labeled data for robustifying BEV segmentation. Yet, our preliminary findings reveal that generation noise in synthetic data compromises efficient BEV model learning. To fully harness the potential of synthetic data from world models, this paper proposes NRSeg, a noise-resilient learning framework for BEV semantic segmentation. Specifically, a Perspective-Geometry Consistency Metric (PGCM) is proposed to quantitatively evaluate the guidance capability of generated data for model learning. This metric originates from the alignment measure between the perspective road mask of generated data and the mask projected from the BEV labels. Moreover, a Bi-Distribution Parallel Prediction (BiDPP) is designed to enhance the inherent robustness of the model, where the learning process is constrained through parallel prediction of multinomial and Dirichlet distributions. The former efficiently predicts semantic probabilities, whereas the latter adopts evidential deep learning to realize uncertainty quantification. Furthermore, a Hierarchical Local Semantic Exclusion (HLSE) module is designed to address the non-mutual exclusivity inherent in BEV semantic segmentation tasks. Experimental results demonstrate that NRSeg achieves state-of-the-art performance, yielding the highest improvements in mIoU of 13.8% and 11.4% in unsupervised and semi-supervised BEV segmentation tasks, respectively. The source code will be made publicly available at https://github.com/lynn-yu/NRSeg.