Unsupervised Monocular Road Segmentation for Autonomous Driving via Scene Geometry

2025-10-19

AI Summary

This paper addresses unsupervised monocular road segmentation by proposing a weakly supervised learning framework that eliminates the need for manual annotations. Methodologically, it jointly leverages monocular geometric priors—namely depth and surface normal estimation—with optical-flow-guided feature point tracking across frames to generate reliable weak labels. To further refine label quality and temporal stability, the framework incorporates mutual information maximization and temporal consistency constraints during joint optimization of the segmentation model. The key contributions are: (i) the first integration of geometric and motion cues for collaborative modeling in unsupervised road segmentation; and (ii) the application of information-theoretic principles to enforce cross-frame prediction consistency. Evaluated on Cityscapes, the method achieves 82.0% IoU—matching the performance of fully supervised baselines—while demonstrating superior temporal robustness.

Technology Category

• Computer Vision: Segmentation
• Machine Learning: Unsupervised & Self-Supervised Learning
• Search and Optimization: Learning to Search

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

This paper presents a fully unsupervised approach for binary road segmentation (road vs. non-road), eliminating the reliance on costly manually labeled datasets. The method leverages scene geometry and temporal cues to distinguish road from non-road regions. Weak labels are first generated from geometric priors, marking pixels above the horizon as non-road and a predefined quadrilateral in front of the vehicle as road. In a refinement stage, temporal consistency is enforced by tracking local feature points across frames and penalizing inconsistent label assignments using mutual information maximization. This enhances both precision and temporal stability. On the Cityscapes dataset, the model achieves an Intersection-over-Union (IoU) of 0.82, demonstrating high accuracy with a simple design. These findings demonstrate the potential of combining geometric constraints and temporal consistency for scalable unsupervised road segmentation in autonomous driving.

Problem

• Unsupervised road segmentation without manual labels
• Leveraging geometry and temporal cues for road detection
• Achieving high accuracy with simple geometric constraints

Innovation

• Unsupervised road segmentation using scene geometry
• Generating weak labels from geometric horizon priors
• Enforcing temporal consistency through feature tracking