Physics-Informed Temporal U-Net for High-Fidelity Fluid Interpolation
arXiv:2604.23372v1 Announce Type: cross
Abstract: Reconstructing high-fidelity fluid dynamics from sparse temporal observations is quite challenging, mainly due to the chaotic and non-linear nature of fluid transport. Standard deep learning-based inte…