Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
HELIOS introduces a novel approach for continuous outdoor urban scene relighting that operates without the need for paired images or synthetic data, leveraging unlabeled real-world datasets. By integrating albedo-based conditioning into a cycle-consistent diffusion pipeline and employing a robust albedo distillation strategy, HELIOS effectively manages low-visibility nighttime conditions while ensuring structural stability. Extensive evaluations reveal that HELIOS achieves superior performance in both night-to-day and day-to-night relighting tasks compared to existing state-of-the-art methods.
HELIOS outperforms traditional relighting techniques by enabling realistic scene illumination transitions without paired training data, transforming how we approach outdoor lighting challenges.
Modifying the illumination of driving images is a fundamental challenge, as most datasets are captured at specific times of day. Existing methods rely on synthetic data or paired multi-illumination supervision, which limits their generalization to the diverse and challenging conditions of real-world scenarios. To address this, we propose HELIOS, a novel image relighting approach that relies on unlabeled real-world datasets without requiring any paired images for training. Our approach integrates albedo-based conditioning into a cycle-consistent diffusion pipeline to prevent identity collapse and ensure accurate domain translation. To handle low-visibility nighttime conditions, we introduce a robust albedo distillation strategy that transfers structural stability from the daytime domain. Additionally, we replace traditional text prompts with a fine-grained control mechanism based on GPS-derived solar angles, enabling smooth and continuous lighting manipulation across the day-night cycle. Through extensive evaluation and a user study, we demonstrate that HELIOS produces structurally consistent and realistic results in both night-to-day and day-to-night tasks, outperforming state-of-the-art methods.