Summary of Luminet: Latent Intrinsics Meets Diffusion Models For Indoor Scene Relighting, by Xiaoyan Xing et al.
LumiNet: Latent Intrinsics Meets Diffusion Models for Indoor Scene Relighting
by Xiaoyan Xing, Konrad Groh, Sezer Karaoglu, Theo Gevers, Anand Bhattad
First submitted to arxiv on: 29 Nov 2024
Categories
- Main: Computer Vision and Pattern Recognition (cs.CV)
- Secondary: Graphics (cs.GR); Machine Learning (cs.LG)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary LumiNet is a novel architecture that uses generative models and latent intrinsic representations to effectively transfer lighting between images. It takes a source image and a target lighting image as input, and generates a relit version of the source scene that captures the target’s lighting. The approach makes two key contributions: a data curation strategy for training, and a modified diffusion-based ControlNet that processes both latent intrinsic properties from the source image and latent extrinsic properties from the target image. Additionally, a learned adaptor (MLP) is used to fine-tune the model through cross-attention. This architecture is demonstrated on various lighting transfer tasks and shows promising results. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary LumiNet is a new way to make pictures look like they were taken in different lighting conditions. It uses special computer models and takes two pictures as input: one of an object or scene, and another of the same thing but with different lighting. The model then creates a new picture that combines the original object or scene with the desired lighting. This is useful for many applications, such as making movies or photographs look like they were taken in a specific time period or location. |
Keywords
» Artificial intelligence » Cross attention » Diffusion
