Zehao Zhu, Yuliang Zou, Chiyu Max Jiang, Bo Sun, Vincent Casser, Xiukun Huang, Jiahao Wang, Zhenpei Yang, Ruiqi Gao, Leonidas Guibas, Mingxing Tan, Dragomir Anguelov
SceneCrafter is a new tool that allows for realistic editing of multi-view driving scenes, addressing challenges in 3D consistency and data generation for autonomous vehicle simulation.
SceneCrafter is a tool designed to help improve the simulation environments used to test self-driving cars. Unlike traditional methods that create scenes from scratch, SceneCrafter edits real driving footage to simulate different conditions like traffic and weather, making the simulations more realistic. This tool ensures that the changes made to one camera view are consistent across all views, which is important for accurate testing. SceneCrafter uses advanced techniques to create realistic scene edits, providing a more reliable way to test autonomous vehicles under various scenarios.
