Learning to Estimate Single-View Volumetric Flow Motions without 3D Supervision

Autor:	Franz, Erik, Solenthaler, Barbara, Thuerey, Nils
Rok vydání:	2023
Předmět:	Computer Science - Computer Vision and Pattern Recognition Computer Science - Machine Learning Physics - Fluid Dynamics
Druh dokumentu:	Working Paper
Popis:	We address the challenging problem of jointly inferring the 3D flow and volumetric densities moving in a fluid from a monocular input video with a deep neural network. Despite the complexity of this task, we show that it is possible to train the corresponding networks without requiring any 3D ground truth for training. In the absence of ground truth data we can train our model with observations from real-world capture setups instead of relying on synthetic reconstructions. We make this unsupervised training approach possible by first generating an initial prototype volume which is then moved and transported over time without the need for volumetric supervision. Our approach relies purely on image-based losses, an adversarial discriminator network, and regularization. Our method can estimate long-term sequences in a stable manner, while achieving closely matching targets for inputs such as rising smoke plumes. Comment: ICLR 2023 poster, source code: https://github.com/tum-pbs/Neural-Global-Transport
Databáze:	arXiv
Externí odkaz:	http://arxiv.org/abs/2302.14470 Zobrazit plný text záznamu View this record from Arxiv