Unsupervised Feature Extraction Using GBDX

Alexis Smith, CEO, IMGeospatial


Automated cloud based analytics change the way you see your world

IMGeospatial has developed proprietary software that has the ability to extract 15 classifications (and counting) of features totally unsupervised w over 95 per cent accuracy, compared to closest competitor returning at best 65-70 per cent accuracy whilst capable of extracting only four basic classifications of terrain (not specific features).

Being able to classify features over such vast geographic areas at these error co-efficients is a real game changer. Alexis Smith, CEO of IMGeospatial, explains how this totally unsupervised multi-feature extraction is a quantum leap for the industry.

About the Speaker

Alexis Smith, CEO, IMGeospatial

Alexis has enjoyed a successful career working in the global business and entertainment sectors in over 50 countries, and is now applying her entrepreneurial skills to the field of automated analytics. Her company IMGeospatial creates AUTOMATICALLY SMARTERâ„¢ products based on machine learning, artificial intelligence, automatic feature extraction, remote sensing and multi-scopic data, providing solutions to a variety of challenges faced by many industry sectors across the globe.

Learn more about unsupervised feature extraction using GBDX

Fill out the form to get in touch with DigitalGlobe