Introduction Video will be available soon.
VISION6 is a new, experimental vision recognition A.I. that can abstract shapes and structures of visual objects.
Contrast to machine-learning, VISION6 doesn't require millions of trainings in order to recognize a single object's common appearance.
Just like a human child learns a look of an object in a moment while reading a picture book,
VISION6 performs visual abstraction of an object and learns its look by watching a few number of drawings.
Once VISION6 learns a look of an object from drawings, it can recognize other instances of the kind, also in photographs or animated video input from camera.