Skip to content

DINCAE (Data-Interpolating Convolutional Auto-Encoder) is a neural network to reconstruct missing data in satellite observations.

License

Notifications You must be signed in to change notification settings

gher-uliege/DINCAE.jl

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Build Status codecov documentation stable documentation dev

Issues Issues Commit GitHub top language

DINCAE.jl

DINCAE (Data-Interpolating Convolutional Auto-Encoder) is a neural network to reconstruct missing data in satellite observations. This repository contains the Julia port of DINCAE. The original Python code is no longer maintained.

Utilities (for plotting and data preparation) are available in a separate repository https://github.com/gher-uliege/DINCAE_utils.jl

The method is described in the following articles:

  • Barth, A., Alvera-Azcárate, A., Licer, M., & Beckers, J.-M. (2020). DINCAE 1.0: a convolutional neural network with error estimates to reconstruct sea surface temperature satellite observations. Geoscientific Model Development, 13(3), 1609–1622. https://doi.org/10.5194/gmd-13-1609-2020
  • Barth, A., Alvera-Azcárate, A., Troupin, C., & Beckers, J.-M. (2022). DINCAE 2.0: multivariate convolutional neural network with error estimates to reconstruct sea surface temperature satellite and altimetry observations. Geoscientific Model Development, 15(5), 2183–2196. https://doi.org/10.5194/gmd-15-2183-2022

(click here for the BibTeX entry).

Panel (a) is the original data where we have added clouds (panel (b)). The reconstuction based on the data in panel (b) is shown in panel (c) together with its expected standard deviation error (panel (d))

DINCAE is intended to be used with a GPU with CUDA support (NVIDIA GPU). The code can also run on a CPU but which will be quite slow.

Installation

You need Julia (version 1.9 or later) to run DINCAE. The command line interface of Julia is sufficient for DINCAE. If you are using Linux (on a x86_64 CPU), installing and running Julia is as easy as running these shell commands:

curl https://julialang-s3.julialang.org/bin/linux/x64/1.10/julia-1.10.2-linux-x86_64.tar.gz | tar -xzf -
julia-1.10.2/bin/julia

This installs Julia in the current directory under the folder julia-1.10.2. For more information, other platforms and the current julia version, please see platform specific instructions for further installation instructions. You can check the latest available version of Julia at https://julialang.org/downloads but avoid beta releases and release candidates if you are new to Julia.

Inside a Julia terminal, you can download and install DINCAE and DINCAE_utils by issuing these commands:

using Pkg
Pkg.add(url="https://github.com/gher-uliege/DINCAE.jl", rev="main")
Pkg.add(url="https://github.com/gher-uliege/DINCAE_utils.jl", rev="main")

CUDA support

To enable (optional) CUDA support on NVIDIA GPUs one need to install also the packages CUDA and cuDNN:

using Pkg 
Pkg.add("CUDA")
Pkg.add("cuDNN")

With some adaptions to DINCAE.jl, one can probably also use AMD GPUs (with the package AMDGPU) and Apple Silicon (with the package Metal). PRs to implement support of these GPUs would be very welcome.

After this, you should be able to load DINCAE with:

using DINCAE

Checking CUDA installation

To confirm that CUDA is functional to use the GPU (otherwise the CPU is used and the code will be much slower), the following command:

CUDA.functional()

should return true.

Updating DINCAE

To update DINCAE, run the following command and restart Julia (or restart the jupyter notebook kernel using Kernel -> Restart):

using Pkg
Pkg.update("DINCAE")

Note that Julia does not directly delete the previous installed version. To check if you have the latest version run the following command:

using Pkg
Pkg.status()

The latest version number is available from here.

Documentation

More information is available in the documentation and the tutorial (available as script and jupyter notebook).

Publications

About the code

  • Barth, A., Alvera-Azcárate, A., Licer, M., & Beckers, J.-M. (2020). DINCAE 1.0: a convolutional neural network with error estimates to reconstruct sea surface temperature satellite observations. Geoscientific Model Development, 13(3), 1609–1622. https://doi.org/10.5194/gmd-13-1609-2020
  • Barth, A., Alvera-Azcárate, A., Troupin, C., & Beckers, J.-M. (2022). DINCAE 2.0: multivariate convolutional neural network with error estimates to reconstruct sea surface temperature satellite and altimetry observations. Geoscientific Model Development, 15(5), 2183–2196. https://doi.org/10.5194/gmd-15-2183-2022

Applications

  • Han, Z., He, Y., Liu, G., & Perrie, W. (2020). Application of DINCAE to Reconstruct the Gaps in Chlorophyll-a Satellite Observations in the South China Sea and West Philippine Sea. Remote Sensing, 12(3), 480. https://doi.org/10.3390/rs12030480
  • Ji, C., Zhang, Y., Cheng, Q., & Tsou, J. Y. (2021). Investigating ocean surface responses to typhoons using reconstructed satellite data. International Journal of Applied Earth Observation and Geoinformation, 103, 102474. https://doi.org/10.1016/j.jag.2021.102474
  • Jung, S., Yoo, C., & Im, J. (2022). High-Resolution Seamless Daily Sea Surface Temperature Based on Satellite Data Fusion and Machine Learning over Kuroshio Extension. Remote Sensing, 14(3), 575. https://doi.org/10.3390/rs14030575
  • Luo, X., Song, J., Guo, J., Fu, Y., Wang, L. & Cai, Y. (2022). Reconstruction of chlorophyll-a satellite data in Bohai and Yellow sea based on DINCAE method International. Journal of Remote Sensing, 43, 3336-3358. https://doi.org/10.1080/01431161.2022.2090872

Thank you for citing relevant previous work in DINCAE if you make a scientific publication. A bibtex entry can be generated from the DOI by using for example curl -LH "Accept: application/x-bibtex" 'https://doi.org/10.5194/gmd-15-2183-2022'.

Feel free to add your publications by making a pull request or opening an issue.

About

DINCAE (Data-Interpolating Convolutional Auto-Encoder) is a neural network to reconstruct missing data in satellite observations.

Topics

Resources

License

Stars

Watchers

Forks

Packages

No packages published

Languages