Perceptual system

A perceptual system is a computational system (biological or artificial) designed to make inferences about properties of a physical environment based on scenes. Other definitions may exist.

In this context, a scene is defined as information that can flow from a physical environment into a computational system via sensory transduction. A sensory organ (biological or artificial) is used to capture this information. Therefore, any perceptual system must incorporate input from at least one sensory organ.

Examples of perceptual systems include:

• The human visual system
• The human auditory system
• A bat's sonar/echolocation system
• A man-made light meter
• A man-made motion detector

Research in the field of perceptual systems focuses on computational aspects of perception. For this reason, there is significant overlap with neuroscience, sensor design, natural scene statistics,^[1][2] and computer science.

References

1. ↑ Lewicki, Michael S. (2002), Efficient coding of natural sounds. Nature Neuroscience 5(4):356-363. PDF
2. ↑ Geisler, William S. (2008), Visual perception and the statistical properties of natural scenes. Annual Review of Psychology 59:167-192. PDF

External links

• Center for Perceptual Systems, University of Texas at Austin