Solar Altitude Guided Scene Illumination

TL;DR

This paper introduces solar altitude as a global conditioning variable for synthetic scene illumination in autonomous driving, enabling realistic daytime lighting without manual labels, and improves diffusion models' accuracy in capturing lighting variations.

Contribution

It proposes using solar altitude as a label-free, globally computable condition for synthetic data generation, addressing daytime variation challenges in scene illumination modeling.

Findings

Solar altitude effectively captures lighting variations in synthetic scenes.

The normalization approach improves diffusion model performance in illumination modeling.

Synthetic data quality is enhanced by the proposed conditioning method.

Abstract

The development of safe and robust autonomous driving functions is heavily dependent on large-scale, high-quality sensor data. However, real-world data acquisition requires extensive human labor and is strongly limited by factors such as labeling cost, driver safety protocols and scenario coverage. Thus, multiple lines of work focus on the conditional generation of synthetic camera sensor data. We identify a significant gap in research regarding daytime variation, presumably caused by the scarcity of available labels. Consequently, we present solar altitude as global conditioning variable. It is readily computable from latitude-longitude coordinates and local time, eliminating the need for manual labeling. Our work is complemented by a tailored normalization approach, targeting the sensitivity of daylight towards small numeric changes in altitude. We demonstrate its ability to accurately capture lighting characteristics and illumination-dependent image noise in the context of diffusion models.