A Caltech Library Service

Seeing the Wind: Visual Wind Speed Prediction with a Coupled Convolutional and Recurrent Neural Network

Cardona, Jennifer L. and Howland, Michael F. and Dabiri, John O. (2019) Seeing the Wind: Visual Wind Speed Prediction with a Coupled Convolutional and Recurrent Neural Network. In: 33rd Conference on Neural Information Processing Systems. Neural Information Processing Systems Foundation, Inc. , Art. No. 9078.

[img] PDF - Published Version
See Usage Policy.

[img] PDF - Submitted Version
See Usage Policy.

[img] Archive (ZIP) - Supplemental Material
See Usage Policy.


Use this Persistent URL to link to this item:


Wind energy resource quantification, air pollution monitoring, and weather forecasting all rely on rapid, accurate measurement of local wind conditions. Visual observations of the effects of wind---the swaying of trees and flapping of flags, for example---encode information regarding local wind conditions that can potentially be leveraged for visual anemometry that is inexpensive and ubiquitous. Here, we demonstrate a coupled convolutional neural network and recurrent neural network architecture that extracts the wind speed encoded in visually recorded flow-structure interactions of a flag and tree in naturally occurring wind. Predictions for wind speeds ranging from 0.75-11 m/s showed agreement with measurements from a cup anemometer on site, with a root-mean-squared error approaching the natural wind speed variability due to atmospheric turbulence. Generalizability of the network was demonstrated by successful prediction of wind speed based on recordings of other flags in the field and in a controlled wind tunnel test. Furthermore, physics-based scaling of the flapping dynamics accurately predicts the dependence of the network performance on the video frame rate and duration.

Item Type:Book Section
Related URLs:
URLURL TypeDescription Paper
Howland, Michael F.0000-0002-2878-3874
Dabiri, John O.0000-0002-6722-9008
Additional Information:© 2019 Neural Information Processing Systems Foundation, Inc. The authors acknowledge Kelyn Wood, who assisted in setup for the wind tunnel test set. J.L.C. is funded through the Brit and Alex d’Arbeloff Stanford Graduate Fellowship, and M.F.H. is funded through a National Science Foundation Graduate Research Fellowship under Grant DGE-1656518 and a Stanford Graduate Fellowship.
Funding AgencyGrant Number
Stanford UniversityUNSPECIFIED
NSF Graduate Research FellowshipDGE-1656518
Record Number:CaltechAUTHORS:20200710-104503016
Persistent URL:
Usage Policy:No commercial reproduction, distribution, display or performance rights in this work are provided.
ID Code:104324
Deposited By: Tony Diaz
Deposited On:10 Jul 2020 18:12
Last Modified:26 Jul 2021 19:26

Repository Staff Only: item control page