Selective Attention Improves Learning

Abstract. We demonstrate that selective attention can improve learning. Considerably fewer samples are needed to learn a source separation problem when the inputs are pre-segmented by the proposed model. The model combines biased-competition model for attention with a habituation mechanism which allows the focus of attention to switch from one object to another. The criteria for segmenting objects are estimated from data and are shown to generalise to new objects. Key words: Selective attention, perceptual learning, segmentation

• Type: Conference
• Year: 2009
• Where: ICANN
• Authors: Antti Yli-Krekola, Jaakko Särelä, Harri Valpola
• Keywords: Artificial Intelligence, ICANN 2009, Model Combines Biased-competition, Selective Attention, Source Separation Problem