CVPR2025
FluidNexus: 3D Fluid Reconstruction and Prediction from a Single Video
Yue Gao, Hong-Xing Yu, Bo Zhu, Jiajun Wu
Abstract
Input single video Future prediction (novel view) Reconstructed fluid Interaction: Wind (novel view) Interaction: Object (novel view) + Wind + Object + Time Figure 1. From a single fluid video (top left), we reconstruct the 3D fluid (middle), predict future evolution (bottom left), and simulate wind-fluid interaction (top right) / object-fluid interactions (bottom right).