A Framework for Continuous Learning of Simple Visual Concepts

We present a continuous learning framework for learning simple visual concepts and its implementation in an artificial cognitive system. The main goal is to learn associations between automatically extracted visual features and words that describe the scene in an open-ended, continuous manner. In particular, we address the problem of crossmodal learning of elementary visual properties and spatial relations; we show that the same learning mechanism can be used to learn both types of concepts. We introduce and analyse several learning modes requiring different levels of tutor supervision, ranging from a completely tutor driven to a completely autonomous exploratory approach.