AI Paper Atlas

Research map / Probabilistic and scientific modeling / Differential equations and PDEs

Physics-Aware Downsampling with Deep Learning for Scalable Flood Modeling

Niv Giladi, Zvika Ben-Haim, Sella Nevo, Yossi Matias, Daniel Soudry

NeurIPS 2021 · Probabilistic and scientific modeling › Differential equations and PDEs · approach: Physics-informed operator learning · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
Replace average pooling with learned DNN downsamplingcoarse-grid flood simulation · Elevation-map generalization testHuber loss: 6.04\cdot10^{-3} -> 3.07\cdot10^{-3}Improved
Evaluate DNN on hardest samples (90th percentile)DNN downsampling · Elevation-map generalization testHuber loss 90 percentile: 2.80\cdot10^{-2} -> 1.10\cdot10^{-2}Improved
Replace average pooling with DNN downsamplingcoarse-grid flood simulation · Elevation-map generalization testInundation error 90 percentile: 5.64\cdot10^{-2} -> 1.77\cdot10^{-2}Improved
Replace average pooling with DNN downsamplingcoarse-grid flood simulation · Elevation-map generalization testMSE: 1.37\cdot10^{-2} -> 6.64\cdot10^{-3}Improved
Replace average pooling with DNN downsamplingcoarse-grid flood simulation · Elevation-map generalization testInundation error: 9.77\cdot10^{-3} -> 4.44\cdot10^{-3}Improved
Train DNN on a single map with multiple boundary conditions vs average poolingDNN downsampling · Single-map multi-boundary-condition testHuber loss vs fine-grid solutionImproved
Train DNN on one hour of flow, evaluate 20 hours on Ganges regionDNN downsampling · Ganges river 175x175 coarse gridflood extent accuracyImproved
Bilateral filter then average pooling instead of plain average pooling (ablation)average pooling baseline · Edge-preserving downsampling comparisonper-sample lossNo clear effect
Bilateral filter then max pooling instead of average pooling (ablation)average pooling baseline · Edge-preserving downsampling comparisonper-sample lossHurt
Train with Huber loss rather than MSE (ablation)DNN training · loss function choiceconvergenceImproved
Use plain ResNet-18 with normalization layers (ablation)DNN architecture · architecture comparisonelevation scale preservationHurt
Remove normalization only in the residual connection (ablation)ResNet-18 · architecture comparisonconvergence and generalizationImproved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy