I recently attended the annual AGU (American Geophysical Union) meeting and, used the time at our booth to talk to many geoscientists about many different topics. One that emerged as a big interest stemming from unfamiliarity is deep learning.
At MathWorks we have been adding to our machine learning and deep learning repertoire ardently. One reflection of this is the newest blog, on deep learning, authored by Steve Eddins, also the author of the Steve on Image Processing blog.
Most applications of deep learning these days are focused on images, with signals coming along strongly now as well. Here's a link to some MATLAB examples/applications in more traditional areas like object detection.
In the sciences, successes have been reported in biological and medical studies.
In domains, such as geosciences, often data can be viewed as an image even if they don't represent an actual picture. We had several discussions at AGU about possible applications of deep learning for seismic problems, atmospheric ones, and oceanographic ones. And deep learning is certainly having some success in some remote sensing applications, from a quick web search.
Do you know of any people or applications currently using deep learning in their work? What domain? What do you see looking forward to the application of deep learning in domains you know most about? Let us know here.
Get the MATLAB code
Published with MATLAB® R2017b
Comments are closed.
2 CommentsOldest to Newest
Give me the steps to identify the faults and recognize the different types of defects in BARE PRINTED CIRCUIT BOARD by computer vision
This FileExchange entry (https://www.mathworks.com/matlabcentral/fileexchange/64070-deep-learning–image-anomaly-detection-for-production-line-~find-scratched-units~ ) contains code that shows how to do anomaly detection using deep learning and is a good place to start.