AD-Relight: Training-Free Banner Relighting via Illumination Translation with Diffusion Priors

TL;DR

AD-Relight is a training-free, multi-stage framework that adapts diffusion-based models at test time to relight custom ad banners seamlessly integrated into scenes, improving realism and user preference.

Contribution

It introduces a novel test-time adaptation method for diffusion models to relight ad banners without additional training, addressing a key challenge in ad personalization.

Findings

AD-Relight outperforms existing relighting baselines and ad-placement methods.

User studies show a preference for AD-Relight outputs over prior approaches.

The framework effectively adapts to new banners without requiring extensive training.

Abstract

The recent surge in content consumption through streaming services has driven a growing demand for personalized content. Personalized advertisements (ads) play a crucial role in enhancing both user engagement and ad effectiveness. A key aspect of ad personalization involves replacing existing regions in a frame with custom, Photoshop-generated banners. However, existing ad-placement pipelines typically rely on simple geometric warping, ignoring the scene's underlying lighting conditions. Similarly, state-of-the-art diffusion-based object insertion and relighting models struggle to accurately relight these newly inserted banners, as they are not trained on ad-banner data, and training such a model for ad banners would require millions of images. This highlights the need for an effective relighting framework that enables seamless integration of custom banners into the original scene. Motivated by this, we present AD-Relight, a novel multi-stage training-free framework that adapts a diffusion-based relighting model at test time to relight newly added Photoshop-generated ad banners. Through extensive evaluation, we demonstrate that AD-Relight outperforms both relighting baselines and existing ad-placement methods based on simple warping. User studies further show that participants consistently prefer the outputs of AD-Relight over those of prior approaches.