|
Research![]()
![]() My research concentrate on high-quality object insertion and lighting estimation, leveraging diffusion models to achieve seamless, realistic object compositing across diverse scenes. My goal is to develop compositing methods that integrate precise shading and shadow with user inputs, and accurate environmental conditions while minimizing the need for extensive paired data or supervised training. |
![]() |
SpotLight: Shadow-Guided Object Relighting via DiffusionFrédéric Fortier-Chouinard, Zitian Zhang, Louis-Etienne Messier, Mathieu Garon, Anand Bhattad, Jean-François Lalonde Arxiv arxiv / website / SpotLight allows precise local lighting control by specifying the desired shadows of the inserted object. This approach accurately reshades the object and properly harmonizes the object with the target background, without any additional training. |
![]() |
ZeroComp: Zero-shot Object Compositing from Image Intrinsics via DiffusionZitian Zhang, Frédéric Fortier-Chouinard, Mathieu Garon, Anand Bhattad, Jean-François Lalonde WACV 2025 (Oral) arxiv / code / poster / website / supplementary / Zero-shot 3D object compositing approach that does not require paired composite-scene images during training and can be easily extended to 2D object compositing and material editing. |
![]() |
Delving into Multi-illumination Monocular Depth Estimation: A New Dataset and MethodYuan Liang, Zitian Zhang, Chuhua Xian, Shengfeng He TMM paper / code / dataset / An indoor RGB-D dataset with multi-illuminations, and a method for consistent depth estimation under changing lighting. |
Other ProjectsThese include intern projects, and other projects that I have worked on. |
Oasis: A survival simulation game (unfinished)Intern_in_Alibaba_Lingxi_Interactive 2022-08-17 In a scorching desert, the player discovers an oasis and must lead people to work and survive there. |
Design and source code from Jon Barron's website and Leonid Keselman's website. |