Learning Feature Integration for Attention Control

Luiz Gonçalves

We present current efforts towards developing a learning approach for the integration of features extracted from multi-modal sensors, used to guide the attention behavior of robotic agents. Basically, a pre-attention mechanism enhances attentional features that are the most relevant to the current task according to a weighting strategy that can be learned. Then, an attention shift mechanism can select one between the various activated stimuli, in order for a robot to foveate on it. The model can be applied in many different situations and different tasks including top-down or bottom-up aspects of attention control. Also, in this approach, we consider the robot moving resources or so to improve the (visual) sensory information. Clique no link abaixo para buscar o texto completo deste trabalho na Web: Buscar na Web

Biblioteca Digital Brasileira de Computação - Contato:
     Mantida por: