Logical Vision: Meta-Interpretive Learning for Simple Geometrical Concepts.
Wang-Zhou Dai, Stephen Muggleton, Zhi‐Hua Zhou · 2015
Abstract. Progress in statistical learning in recent years has enabled comput-ers to recognize objects with near-human ability. However, recent studies have revealed particular drawbacks in current computer vision systems which sug-gest there exist considerable differences between the way these systems function compared with human visual cognition. Major differences are that: 1) current computer vision systems learn high-level notions directly from the low-level fea-ture space, which makes them sensitive to low-level characteristics changing. 2) typical computer vision systems learn visual concepts discriminatively instead of encoding the knowledge necessary to produce a visual representation of the class. In this paper, we introduce a framework referred as Logical Vision which is demonstrated on learning visual concepts constructively and symbolically. It first constructively extracts logical facts of mid-level features, then generative Meta-Interpretive Learning technique is applied to learn high-level notions because it is capable of learning recursions, inventing predicates and so on. Owing to its symbolic representation paradigm, in our implementation, Logical Vision is fully implemented in Prolog apart from low-level image feature extraction primitives. Experiments are conducted on learning shapes (e.g. triangles, quadrilaterals, etc.), regular polygons and right-angle triangles. These demonstrates that learning vi-sual concepts constructively and symbolically is effective. 1