F-RNG: Feed-Forward Relightable Neural Gaussians
2026-05-25 • Graphics
GraphicsComputer Vision and Pattern Recognition
AI summaryⓘ
The authors present F-RNG, a new method that can create 3D models from a few pictures which can be realistically relit under different lighting conditions. Unlike previous methods that need many images or retraining for new lights, F-RNG uses a clever combination of existing models and learned lighting information to make relightable 3D assets quickly and efficiently. This approach results in faster relighting—about 25 times quicker—and better visual quality compared to previous techniques. Their method also doesn’t require big computational resources and can improve as better base models become available.
3D Gaussian SplattingRelightable 3D AssetsSparse-View ReconstructionFeed-Forward ModelsIntrinsic DecompositionNeural RenderingLatent-Interpolated GeometryAppearance DistillationScene OptimizationLight Conditioning
Authors
Guangming Fu, Jiahui Fan, Jian Yang, Miloš Hašan, Beibei Wang
Abstract
Capturing relightable 3D assets from real-world objects is a widely researched problem. Several per-scene optimization-based methods, based on 3D Gaussian splatting (3DGS), support relighting; however, they usually require dense input views, and their overfitting nature makes it difficult to generalize across scenes. Unlike per-scene optimization methods, generalized feed-forward models can directly reconstruct Gaussians from sparse input views. However, the resulting assets have baked-in illumination and cannot be easily used for relighting. In this paper, we present F-RNG, a feed-forward framework that directly generates relightable 3DGS assets from sparse-view inputs. Training such a model from scratch can require massive data and computing resources, and it is especially challenging to generate relightable assets in a feed-forward manner with acceptable cost. We develop F-RNG upon an existing large reconstruction model (LRM) to extract relightable representations, while also utilizing priors from an intrinsic decomposition model (IDM). Specifically, we first introduce a latent-interpolated fine-grained geometry synthesis to enhance the LRM's geometry representation. Second, we propose a prior-guided relightable appearance distillation to extract relightable neural representations by incorporating IDM priors. Finally, a universal neural renderer enables flexible and high-fidelity relighting. F-RNG requires neither re-training nor fine-tuning of the underlying LRMs, thus can automatically benefit from better LRMs and IDMs in the future. With only small networks that can be trained with affordable data and computational resources, F-RNG avoids the repetitive inference of large models under different light conditions. By comparison to the state-of-the-art LRM-based relighting method, F-RNG achieves ~25x faster relighting, as well as superior quality (~+2.0 dB).