Download the Easy! source code or a binary installer.
June 24, 2013: The CVPR Tutorial was well attended and with an exciting conference ahead, we will try to generate a lot of buzz for Easy!. The technical document that was in such high demand is now available for download: Easy.pdf.
June 11, 2013: The stable master branch was updated to CVAC Version 0.3 with lots of new features. Also available now are binary distribution packages for various platforms.
April 30, 2013: Lots of activity on the "devel" branch: Amongst others additions, there is now a hookup to the popular LabelMe tool for annotations. Most work on the python interface is done on this branch, too.
March 7, 2013: We just switched our versioning system to git: find and "clone" the first public release on GitHub
Easy! Computer Vision ...
- ... is a framework for anybody who wants to tap into the power for computer vision quickly and easily: application developers that would like to use computer vision, computer vision algorithm developers, contracting officers, and program managers.
- ... connects algorithms with data sets, e.g. from the Computer Vision Algorithm Collection (CVAC) or from OpenCV and from the Caltech101 corpus or LabelMe.
- ... enables transparent remote service invocation, cross-language communication, and a plug-and-play approach to model training, testing, and deployment.
Here is a code sample that uses a pre-trained model for vehicle detection in the SampleImage.jpg:
detector = easy.getDetector( "bowTest:default -p 10004" ) results = easy.detect( detector, "VehicleModel.zip", "SampleImage.jpg" ) easy.printResults( results )
To train a model for vehicle detection using a "Bag of Words" method, you would write:
categories = easy.getDataSet( "Caltech101.properties" ) runset = easy.createRunSet( categories["car_side"] ) trainer = easy.getTrainer( "bowTrain:default -p 10003" ) VehicleModel = easy.train( trainer, runset )
Further information about Easy! Computer Vision:Learn how to make Easy! work for you at the CVPR 2013 Tutorial.
Read more about Easy! in the user documentation, the API documentation, read the source code and program output of a growing number of mini tutorials, read answers to some of the most Frequently Asked Questions (FAQ), read details about the data exchange, take a look at the US Government repository of vision algorithms at Forge.mil (CAC Card required for access), or just download Easy! and start using it!