Zhang, Xin (National University of Defense Technology), Sun, Zhenping (National University of Defense Technology), Bu, Yafeng (National University of Defense Technology), Zeng, Jun (National University of Defense Technology), Li, Xiaohui (National University of Defense Technology)

Trajectory-Guided Self-Supervised Traversability Analysis for Off-Road Navigation under Low-Light Conditions

Scheduled for presentation during the Video Session "On-Demand Video Presentations" (VP-VP), Saturday, November 22, 2025, 08:00−18:00, On-Demand Platform

2025 IEEE 28th International Conference on Intelligent Transportation Systems (ITSC), November 18-21, 2025, Gold Coast, Australia

This information is tentative and subject to change. Compiled on April 2, 2026

Abstract

To mitigate the significant degradation in traversability analysis accuracy caused by the limited perception capabilities of visible cameras in nighttime environments, we propose a trajectory-guided self-supervised traversability analysis method utilizing infrared camera. This approach leverages online adaptive clustering of high-dimensional features derived from historical trajectories and computes cosine similarity by integrating these features with those extracted from infrared images. This enables robust traversability analysis even under low-light or no-light conditions. Experimental results demonstrate that the proposed method effectively addresses the challenges of feature representation in low-light environments, significantly enhancing the accuracy of traversability analysis in nighttime scenarios.