Corpus ID: 211171652

Learning Similarity Metrics for Numerical Simulations

@inproceedings{Kohl2020LearningSM,
  title={Learning Similarity Metrics for Numerical Simulations},
  author={Georg Kohl and Kiwon Um and N. Thuerey},
  booktitle={ICML},
  year={2020}
}
We propose a neural network-based approach that computes a stable and generalizing metric (LSiM) to compare data from a variety of numerical simulation sources. We focus on scalar time-dependent 2D data that commonly arises from motion and transport-based partial differential equations (PDEs). Our method employs a Siamese network architecture that is motivated by the mathematical properties of a metric. We leverage a controllable data generation setup with PDE solvers to create increasingly… Expand

References

SHOWING 1-10 OF 80 REFERENCES
Imagenet classifica­ tion with deep convolutional neural networks
  • Communications of the ACM,
  • 2017
Ima­ genet classification with deep convolutional neural net­ works
  • Communications of the ACM,
  • 2017
ImageNet classification with deep convolutional neural networks
  • 62,892
  • Highly Influential
  • PDF
Data exploration of turbulence simulations using a database cluster
  • 175
  • Highly Influential
  • PDF
Spot the Dif­ ference: Accuracy of Numerical Simulations via the Hu­ man
  • Visual System. CoRR,
  • 2019
The Unreasonable Effectiveness of Deep Features as a Perceptual Metric
  • 1,400
  • Highly Influential
  • PDF
Image quality assessment: From error visibility to struc­ tural similarity
  • IEEE Transactions on Image Processing,
  • 2004
Image quality assessment: from error visibility to structural similarity
  • 26,418
  • Highly Influential
  • PDF
WeatherBench: A benchmark dataset for data-driven weather forecasting
  • 39
  • PDF
...
1
2
3
4
5
...