Abstract

We present a simple yet effective technique to estimate lighting in a single input image. Current techniques rely heavily on HDR panorama datasets to train neural networks to regress an input with limited field-of-view to a full environment map. However, these approaches often struggle with real-world, uncontrolled settings due to the limited diversity and size of their datasets. To address this problem, we leverage diffusion models trained on billions of standard images to render a chrome ball into the input image. Despite its simplicity, this task remains challenging: the diffusion models often insert incorrect or inconsistent objects and cannot readily generate images in HDR format. Our research uncovers a surprising relationship between the appearance of chrome balls and the initial diffusion noise map, which we utilize to consistently generate high-quality chrome balls. We further fine-tune an LDR diffusion model (Stable Diffusion XL) with LoRA, enabling it to perform exposure bracketing for HDR light estimation. Our method produces convincing light estimates across diverse settings and demonstrates superior generalization to in-the-wild scenarios.

Paper: https://arxiv.org/abs/2312.09168

Code: https://github.com/DiffusionLight/DiffusionLight

Colab Notebook: https://colab.research.google.com/drive/15pC4qb9mEtRYsW3utXkk-jnaeVxUy-0S?usp=sharing&sandboxMode=true

Project Page:https://diffusionlight.github.io/

HuggingFace model card: https://huggingface.co/DiffusionLight/DiffusionLight

Score measurement: https://vistec-my.sharepoint.com/:f:/g/personal/pakkapon_p_s19_vistec_ac_th/EvBHbnLrVnZArhQTcboh6qkBGcSqUqzdgx13iZ2IsLPzOw?e=D9lSPq