During evolution nature has developed a rich variety of senses that perceive and process physical signals coming from diverse environments. Here we derive a general mathematical framework predicting performance and neuronal characteristics of arbitrary sensory systems. We substantiate this work by treating three modalities as concrete examples. First, we show how optimality relates to echo suppression in nature and technology. Second, we explore the capabilities and limitations of the lateral-line system to localize and simultaneously recognize hydrodynamic objects. Third, we resolve the paradox of a snake's infrared sense converting low resolution to a distinct neuronal representation. Finally, we introduce, analyze, and simulate the neuronal concept of integrated Multimodal Teaching (iMT) that, consistently with and even extending the usual experimental interpretation, explains how an intrinsic alignment of sensory maps can develop.
«
During evolution nature has developed a rich variety of senses that perceive and process physical signals coming from diverse environments. Here we derive a general mathematical framework predicting performance and neuronal characteristics of arbitrary sensory systems. We substantiate this work by treating three modalities as concrete examples. First, we show how optimality relates to echo suppression in nature and technology. Second, we explore the capabilities and limitations of the lateral-li...
»