We describe a novel method for acquisition of higher-level visual concepts using GP-based learners that process attributed visual primitives derived from raw raster images. The approach uses an original evaluation scheme: individuals-learners are rewarded for being able to restore the essential features (here: shape) of the visual stimulus. The approach is general and does not require any a priori knowledge about the particular application or target concept to be learned; the only prerequisite is universal knowledge related to interpretation of visual information, encoded in nodes of GP trees. The paper demonstrates the performance of the method on a specific visual task of acquiring the concept of a triangle from examples given in a form of raw raster images.
JavaScript jest wyłączony w Twojej przeglądarce internetowej. Włącz go, a następnie odśwież stronę, aby móc w pełni z niej korzystać.