Do animals recognize pictures as representations of 3D objects

A picture is something in which objects can be recognized, even though the objects themselves are not actually present when it is observed. For observers to recognize objects in a picture, visual inputs have to be matched to the representations of the visual objects in memory. Patients who suffer from visual agnosia often exhibit greater impairments in recognizing line-drawings and photographs than real objects, while leaving abilities of perceiving 2D features intact. Hiraoka, Suzuki, Hirayama, and Mori (2009) reported a patient who does not show apparent impairment with real objects and even their photographs, but has difficulties in recognizing line-drawings and silhouettes (i.e., the stimuli are missing much of 2D features of the real objects). In contrast, Turnbull, Driver, and McCarthy (2004) reported the performance of a patient on a variety of experimental tasks that investigated the patient’s ability to extract 2D and 3D information from 2D stimuli. His performance was relatively intact on the 2D tasks (e.g., picture-matching across rotations in the picture plane), but was greatly impaired on the 3D tasks (mental rotation in depth, relative depth judgments within line-drawings and shaded photographs, and the discrimination of line-drawings depicting “possible” and “impossible” 3D objects). The finding suggested that his difficulties in recognizing visually presented objects was due to the deficit in deriving 3D structure from the pictorial depth-cues within 2D images. It has also been reported that patients with agnosic deficits often show a greater impairment in the recognition of visually presented living things (e.g., animals) as compared to non-living things (e.g., tools). Wolk, Coslett, and Glosser (2005) argued that such category-specific visual agnosia suggests that information about an object’s form, as well as sensory-motor information specifying the manner of manipulation, may contribute to object recognition. Generally, visual agnosia is assumed to be derived from impairments in the processes by which visual information contacts stored knowledge of objects or representations of 3D objects in memory.