From Jonathan Warrick:


Dear colleagues,

This summer and fall, as part of the USGS Community for Data Integration (CDI) funded project, "Mapping land-use, hazard vulnerability and habitat suitability using deep neural networks," we will be developing software tools and data sets for USGS scientists to apply deep machine learning tools to classification tasks on large data sets consisting of images of natural environments.

As part of this effort, we are pleased to announce that we will be holding two 3-day training events.  The first will be at the USGS Southwest Biological Science Center in Flagstaff, AZ, 10 through 12 JulyThere will be a second 3-day, 30-person training event at the USGS National Training Center in Denver, CO, 25 through 27 September

Registration is free.  Attendees will need to make their own travel and accommodation arrangements.  We will be limited to just 30 attendees in each training, and we will be giving priority to USGS and other DOI employees.  We also seek a diverse representation of USGS Mission Areas, Centers and scientists.  So, please sign up soon.

Our workshops will focus on application of deep neural networks to classification tasks using images of natural environments (including oblique, satellite, and UAS imagery), the python programming language, and the tensorflow machine learning framework. Many data sets will be provided, but attendees will also be encouraged to bring their own data. No prior experience of machine learning or python will be strictly required, but this workshop will most suit scientists engaged in or familiar with using imagery (in its broadest definition - short or long-range; mono-, multi- or hyper-spectral) for purposes of classification, mapping (e.g. land cover, habitats, hazards), photogrammetry, geospatial analysis, etc.

For more details on the objectives, scope, and deliverables of the project, including how to sign up for one of the free 3-day training event, and how to sign up to our mailing list to be keep informed on developments, please see our website

More specific details on the course content will appear soon as the details are finalized, but in the meantime, please email us with any questions or comments.


Sincerely,


Daniel Buscombe, daniel.buscombe@nau.edu
Jon Warrick, jwarrick@usgs.gov
Jenna Brown, jennabrown@usgs.gov
Chris Sherwood, csherwood@usgs.gov
Paul Grams, pgrams@usgs.gov