Lecture 3 Visual Perception of Objects

imissyou419's version from 2017-10-27 15:22


Question Answer
What is the mirroring effect and why does it occur?Mirroring at V1/V2 border and V2/V3 border. "like" cells like to be near each other reducing the length of axons and of the wiring of the brain. Suggested during brain's development axons act to pull areas that are heavily connected together forming cortical folds between V1/V2
When does the assembly of lines/edges into objects occur, how does it occur?V2 and V3; Before perceiving that 4 of the 5 lines belong to the box, all V1 cells activated asynchronously. After binding in V2/V3, 4 of the simple cells are grouped within box and fire synchronously. This grouping is fed back to V1 producing synchronous and therefore larger activity; which features are bound is indicated by synchronous activity in cells that encode these features by using extensive reciprocal interconnections
What does the visual system use to group features into objects aka binding?common color, motion, form (nearness of lines or shapes), past experiences
Illusionary contourCells in V2 and some in V1, are activated by both a real contour and an illusory contour; higher areas use this type of cell to bind lines separated by gaps
How does past memories influence what we see?predictive influence; e.g. misspelled example, reading jumbled words
From V3, information diverges to how many higher order visual areas?over 3 dozen (36+)
What happens at higher order areas beyond V3?each of these areas processes some special aspect of what we see. Each area looks at a different attribute of the same movie: motion, colour, etc.
Dorsal streamalong the intra parietal sulcus to posterior parietal cortex, concerned with largely unconscious directing actions (sacchades, grasps, reaches, feedings) to the spatial location of objects (where stream) - activation of these areas direct one's attention to locations, but the selection of approperiate effectors (which area to which) is unconscious
Gets its input from large ganglion cells (magnocellular) in the peripheral retina
Ventral streamalong the inferior temporal lobe, concerned with perception and recognition of objects e.g. faces (what stream)
Gets its input from small ganglion cells (parvocellular) in fovea
Object perception start to end1. V1 - extract simple features e.g. lines, activate pinwheels of same orientation;
2. V2/V3 - features that share common cues, such as lines of similar orientations, are bound together;
3. LOC - combines object parts seen in the contralateral visual field but not, as yet, those in ipsilateral visual field. Elements of objects are extracted from the background on the bases of features that are bound by common color, motion, or form. Codes that something is an object part;
4. FFA in IT - brings together left and right sides and object is recognized.
IT: center for object perception, where cells respond to a particular combination of complex features, that define a particular object, e.g. face. stores the memories of a variety of objects e.g. animals. FFA in IT represent faces, all neurons respond preferentially to faces and a particular face is stored by a cluster of highly selective neurons. A similar visual representation may hold for all objects.
Lesion in LOCvisual agnosia: the inability to perceive all objects through vision. Bilateral lesion of LOC result in inability to recognize any object incl. faces. Difficulty seeing the mouse among the other objects, but when directed to it will reach for the mouse with the correct orientation without knowing it is a mouse.
Lesion in ITvisual agnosia of a particular class of objects e.g. rhino-agnosia. Lesion in FFA lead to Prosopagnosia: specific loss of face recognition (cannot recognize friends or themselves) but can recognize them through voice or gait. Visual acuity and recognition of colors and movements not impaired. Patients can recognize that a face is a face and features such as eye brows, lips, but cannot recognize a particular combination belong to a particular person
Lesion in intraparietal sulcusdifficulty in pointing or grasping accurately e.g. difficulty grasping a computer mouse with the correct orientation
In areas of IT that include FFA1. cells respond selectively to a particular class of objects, e.g. faces, body parts, animals. Cells in some regions of IT respond more to shape of hands than faces or animals. Within each region, some cells are tuned to a particular instances of object e.g. a particular animal.
2. cells exhibit perceptual constancy. Their response is the same independent of a) location of object's image on retina because these cells have large bilateral receptive fields, b) size of the image, c) cue that defines the object's shape (e.g. line, colour, texture, motion).
The area involved in the perception of a particular object is also involved in storing its associated memories.
Meaningability to identify objects that are the same despite appearing from countless viewpoints
What kind of organization does IT have?columnar organization; cells within a column are activated by the same object. 3 neighboring columns respond best to 3 different views of a face or similar letter-like features
Why does the eye scan when it examines an object like a faceyou can only see clearly with the fovea (central 2 degrees of the retina) so you have to scan the face with saccades (point the fovea to each important feature)
Left side specializes inlanguage, i.e. recognition of words and sentences
Right side specializes inobjects with spatially organized features e.g. faces