Comment by naasking

One interaction that captures a multidimensional, multisensory set of perceptions. In an ML training set, say for visual recognition, this would consist at least of hundreds of images from many angles, in different poses and varied lighting.