GSAT: Geometric Traversability Estimation using Self-supervised Learning with Anomaly Detection for Diverse Terrains

2026-03-08

AI Summary

Existing autonomous navigation methods for terrain traversability estimation often rely on handcrafted thresholds or learn solely from positive samples, which limits their generalization capability. This work proposes a self-supervised approach that eliminates the need for negative samples or additional prototypes by constructing a hypersphere in the latent space around positive samples, thereby jointly optimizing anomaly detection and traversability prediction. By circumventing the dependency on labeled data inherent in traditional positive-unlabeled (PU) learning, the method demonstrates robust performance across diverse real-world robotic platforms and simulation environments. Experimental results show significant improvements in navigation reliability and robustness over complex terrains, highlighting the effectiveness of the proposed framework in enhancing autonomous navigation under challenging conditions.

Technology Category

• Intelligent Robots: Localization, Mapping, and Navigation
• Machine Learning: Unsupervised & Self-Supervised Learning
• Search and Optimization: Learning to Search

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Web Mining and Content Analysis: Robustness and generalizability of Web mining methods

Abstract

Safe autonomous navigation requires reliable estimation of environmental traversability. Traditional methods have relied on semantic or geometry-based approaches with human-defined thresholds, but these methods often yield unreliable predictions due to the inherent subjectivity of human supervision. While self-supervised approaches enable robots to learn from their own experience, they still face a fundamental challenge: the positive-only learning problem. To address these limitations, recent studies have employed Positive-Unlabeled (PU) learning, where the core challenge is identifying positive samples without explicit negative supervision. In this work, we propose GSAT, which addresses these limitations by constructing a positive hypersphere in latent space to classify traversable regions through anomaly detection without requiring additional prototypes (e.g., unlabeled or negative). Furthermore, our approach employs joint learning of anomaly classification and traversability prediction to more efficiently utilize robot experience. We comprehensively evaluate the proposed framework through ablation studies, validation on heterogeneous real-world robotic platforms, and autonomous navigation demonstrations in simulation environments.

Problem

• traversability estimation
• self-supervised learning
• anomaly detection
• positive-unlabeled learning
• autonomous navigation

Innovation

• self-supervised learning
• anomaly detection
• traversability estimation
• positive-unlabeled learning
• geometric navigation