Abstract
The Human Sensory System consists of seven distinct elements: the visual, olfactory, gustatory, auditory, superficial somatosensory, inner somatosensory along with vestibular, and thermosensory systems. Each of them is provided with a specialized sets of receptor cells and is able to perceive a stimulus of specific modality. The power to finely discriminate different stimuli of the same modality derives from the properties of the receptor cells which play like fuzzy sets. A stimulus belongs to the receptor fuzzy sets at different degrees. The information of stimuli is encoded in fuzzy information, i.e. in values of membership functions. Due to the structural constraints of sensory systems and the neural noise, human perception is fuzzy-granular and a process of Bayesian probabilistic inference. A link between the Bayes theory of perception and fuzzy sensory information is established in this work. The ideas presented herein will inspire the design of new Artificial Sensory Systems having the purpose of solving Pattern Recognition problems.