ScalarFlow

Abstract

Fig. 1. An example sequence of our ScalarFlow data set. Four frames of our data set are re-rendered as thick smoke. Insets of the corresponding frames of one of the captured video streams, i.e. the real-world reference, are shown in the lower left corners. In this paper, we present ScalarFlow, a first large-scale data set of reconstructions of real-world smoke plumes. In addition, we propose a framework for accurate physics-based reconstructions from a small number of video streams. Central components of our framework are a novel estimation of unseen inflow regions and an efficient optimization scheme constrained by a simulation to capture real-world fluids. Our data set includes a large number of complex natural buoyancy-driven flows. The flows transition to turbulence and contain observable scalar transport processes. As such, the ScalarFlow data set is tailored towards computer graphics, vision, and learning applications. The published data set contains volumetric reconstructions of velocity and density as well as the corresponding input image sequences with calibration data, code, and instructions how to reproduce the commodity hardware capture setup. We further demonstrate one of the many potential applications: a first perceptual evaluation study, which reveals that the complexity of the reconstructed flows would require large simulation resolutions for regular solvers in order to recreate at least parts of the natural complexity contained in the captured data.

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

“…We evaluate our method on both synthetic smoke flows and the real-world captures from the ScalarFlow dataset (Eckert et al, 2019). As a representative of simpler network architectures, i.e.…”

“…This combination is denoted by (RapidGen). We further compare to the direct optimization algorithms of Eckert et al (2019) (ScalarFlow) and Franz et al (2021) (GlobTrans). These can achieve better target matching and long-term transport because each scene is optimized over the full trajectory individually, but this comes at the cost of vastly increased (more than ×100) reconstruction times.…”

“…Fluid flow Fluid flows are traditionally extremely difficult to capture and methods ranging from Schlieren imaging Dalziel et al (2000); Atcheson et al (2008;2009) and particle imaging velocimetry (PIV) methods Grant (1997) 2013) approaches all require specialized setups. The use of multiple commodity cameras simplifies the acquisition (Gregson et al, 2014;Eckert et al, 2019) and allows for view-interpolation to create additional constrains (Zang et al, 2020). Few works have attempted to solve monocular flow estimation in the fluids setting.…”

“…So far, most research is focused on the reconstruction of single scenes. Classic methods use an optimization process working with an explicit volumetric representation (Eckert et al, 2019;Zang et al, 2020;Franz et al, 2021) while some more recent approaches optimize single scenes with neural fields (Mildenhall et al, 2020;Chu et al, 2022). As such an optimization is typically extremely costly, and has to be redone for each new scene, training a neural network to infer an estimate of the motion in a single pass is very appealing.…”

Functional optimization of fluidic devices with differentiable stokes flow

“…Our pipeline for the optimization of fluidic devices has some degree of thematic overlap with fluid control [McNamara et al 2004;Pan et al 2013;Raveendran et al 2012], which has been used broadly in animation applications. In particular, our method shares similarities with Eckert et al [2019] which optimizes external forces in a fluid system in order to match its behavior to real-world thick smoke. Our focus differs from these prior fluid control papers in that we consider the effect of the geometry of the fluid container as the sole factor influencing the resulting flow.…”