Currently we are reworking our GPUVision framework from the lessons learned through the project implementation. This includes some name changes, standarization, and a few very useful updates. More long term goals are: Organization and Ease of use Enhancements Video Support Integration into Photoshop -- GPU Filter for Photoshop(tm) More Debugging methods -- print data range, pixel value, assertions Render to multi-texture support More abstraction from OpenGL and CG New algorithmic More and different packing of data. Possibly packing color channels into 3 areas for normal filtering Feature based, and dense Optical Flow Simple and Complex Neural Networks on the GPU Learning, Running, and Continuous Learning SVM Face Recognition   Ikkjin Ahn, Michael Lehr http://gpuvision.sourceforg.net/ gpuvision@gmail.com July 1, 2005