Precision-Aware Illumination-Disentangled Vision Transformer for Spacecraft 6D Pose Estimation

2026-06-09

AI Summary

Monocular vision-based six-degree-of-freedom pose estimation of spacecraft remains challenging under complex lighting conditions, including specular reflections, shadows, and texture-poor regions. To address this, this work proposes PAID-ViT, a novel Vision Transformer-based architecture that explicitly decouples structural and appearance features. The model incorporates a reliability-aware token aggregation mechanism, foreground mask supervision, and a parameter-free geometric recovery module. By employing a continuous 6D rotation representation, the proposed method significantly reduces translation error and enhances pose estimation robustness under challenging illumination, as demonstrated on the SPEED+ V2 sunlamp benchmark.

Abstract

Vision sensors provide a lightweight solution for spacecraft proximity operations, but monocular spacecraft 6D pose estimation remains difficult under illumination variation, specular reflection, shadowing, weak texture, and background interference. These factors make local visual evidence spatially unreliable and can destabilize pose regression. This article proposes a Precision-Aware Illumination-Disentangled Vision Transformer (PAID-ViT) for robust spacecraft pose estimation.The proposed model separates pose-relevant structure tokens from illumination-sensitive appearance tokens, estimates patch reliability before pose aggregation, and uses foreground mask supervision to preserve silhouette cues. A parameter-free geometric recovery module converts normalized crop coordinates, log-depth, and a continuous 6D rotation representation into camera-frame rotation and translation. Experiments on SPEED+ V2, the SPEED+ validation/lightbox/sunlamp evaluation configuration used in this study, suggest that PAID-ViT reduces translation error and improves robustness in the challenging sunlamp domain, while ablation studies support the complementary roles of illumination disentanglement, reliability-aware token aggregation, mask supervision, and training-side regularization.

Problem

• 6D pose estimation
• illumination variation
• monocular vision
• spacecraft
• visual reliability

Innovation

• illumination disentanglement
• Vision Transformer
• 6D pose estimation
• reliability-aware aggregation
• geometric recovery