摘要: Visual and speech perception tasks, which can be performed with no apparent effort by people, have proved to difficult for machines. This may in part due the absence of cognitive models type proposed above Jakobson. In this paper we attempt give a unified view research machine vision hope that clear appreciation similarities differences lead better information processing perception. Being active both computer speech, found it useful look at problems arisen one domain anticipate corresponding other (Reddy, 1969). Thus, represents comparitive study issues, systems unsolved are, present, interest visual recognition research.