CVPR2025

LeviTor: 3D Trajectory Oriented Image-to-Video Synthesis

Hanlin Wang, Hao Ouyang, Qiuyu Wang, Wen Wang, Ka Leong Cheng, Qifeng Chen, Yujun Shen, Limin Wang

摘要

Near Far Figure 1 . LeviTor is capable of generating videos with controlled occlusion, better depth changes, and complex 3D orbiting movement based on user inputs. Given an initial frame, users can easily draw 3D trajectory using our inference pipeline to represent their desired movements for designated area. We highly recommend viewing the supplementary materials for detailed video demonstrations.