PixHt-Lab: Pixel Height Based Light Effect Generation for Image Compositing
Yichen Sheng, Jianming Zhang, Julien Philip, Yannick Hold-Geoffroy, Xin Sun, He Zhang, Lu Ling, Bedrich Benes
Abstract
a) cutout (b) SSN [48] (c) SSG [47] (d) PixHt-Lab (ours) Figure 1. PixHt-Lab renders realistic reflection and soft shadows on general shadow receivers for 2D cutouts. (a) shows the object cutout and composition background. (b) SSN [48] cannot render soft shadows on walls due to its ground plane assumption. (c) SSG [47] renders specular reflection, but the shadow on the wall is uniformly softened. (d) Our PixHt-Lab renders realistic soft shadows on the wall guided by 3D-aware buffer channels. The shadow is softened according to the background geometry with more realistic details. PixHt-Lab also renders realistic reflections with physically-based surface materials.
Ask about this paper
Your agent reads all of it.
Lune indexed this paper to the last equation, along with the top-tier papers that cite it. Ask a question and the answer quotes them.
Your agent calls
Luneget_paper_fulltext
Free to start. No credit card required.
Terminal
Install the CLIlune papers fulltext 79958a4c-4c81-40d5-aadd-ab0c20d363abCited by top-tier papers13
- GlobalMapper: Arbitrary-Shaped Urban Layout GenerationLiu He, Daniel G. AliagaICCV 2023 · 22 citations
- IMPRINT: Generative Object Compositing by Learning Identity-Preserving RepresentationYizhi Song, Zhifei Zhang, Zhe Lin, Scott Cohen et al.CVPR 2024 · 20 citations
- Shadow Generation for Composite Image Using Diffusion ModelQingyang Liu, Junqi You, Jianting Wang, Xinhao Tao et al.CVPR 2024 · 10 citations
- Dr.Bokeh: DiffeRentiable Occlusion-Aware Bokeh RenderingYichen Sheng, Zixun Yu, Lu Ling, Zhiwen Cao et al.CVPR 2024 · 9 citations
- Refracting Reality: Generating Images with Realistic Transparent ObjectsYue Yin, Enze Tao, Dylan CampbellCVPR 2026 · 2 citations
Builds on13
- PIFu: Pixel-Aligned Implicit Function for High-Resolution Clothed Human DigitizationShunsuke Saito, Zeng Huang, Ryota Natsume, Shigeo Morishima et al.ICCV 2019 · 1,411 citations
- DeepHuman: 3D Human Reconstruction From a Single ImageZerong Zheng, Tao Yu, Yixuan Wei, Qionghai Dai et al.ICCV 2019 · 367 citations
- Mask-ShadowGAN: Learning to Remove Shadows From Unpaired DataXiaowei Hu, Yitong Jiang, Chi-Wing Fu, Pheng-Ann HengICCV 2019 · 255 citations
- Deep Single-Image Portrait RelightingHao Zhou, Sunil Hadap, Kalyan Sunkavalli, David JacobsICCV 2019 · 247 citations
- DenseRaC: Joint 3D Pose and Shape Estimation by Dense Render-and-CompareYuanlu Xu, Song-Chun Zhu, Tony TungICCV 2019 · 204 citations
Related papers
- SSN: Soft Shadow Network for Image CompositingYichen Sheng, Jianming Zhang, Bedrich BenesCVPR 2021
- Ref-GS: Directional Factorization for 2D Gaussian SplattingYoujia Zhang, Anpei Chen, Yumin Wan, Zikai Song et al.CVPR 2025
- Floating No More: Object-Ground Reconstruction from a Single ImageYunze Man, Yichen Sheng, Jianming Zhang, Liang-Yan Gui et al.CVPR 2025
- SpecTRe-GS: Modeling Highly Specular Surfaces with Reflected Nearby Objects by Tracing Rays in 3D Gaussian SplattingJiajun Tang, Fan Fei, Zhihao Li, Xiao Tang et al.CVPR 2025
- EnvGS: Modeling View-Dependent Appearance with Environment GaussianTao Xie, Xi Chen, Zhen Xu, Yiman Xie et al.CVPR 2025
