Final Notes Flashcards
What 2 things can visual angel tell you about?
1) size of object
2) depth
(3) motion based cues
1) relative motion/motion parallax
2) accretion/deletion
3) motion in depth
Motion parallax
- stuff closer to you that fization point looks like it’s moving backwards
thing behing the fixation point are oging in same direction as you
accretion/deletion
accretion - gorwing/uncovering
deletion - being covered up
motion in depth
as something comes closer to you it gets better = looming vs. something that gets smaller as it becomes further away
- innate in babies
Binocular depth cues (sterodepth)
- L and R eyes see things diff. for bincoular disapirty
- this depth perception is really important
Steroscope
- 2 flat images become a 3d images so they look like diff. messages
depth information derived from stereposis vis 2 stages
1) solving the stero correspondance problem
2) calculating the retinal disparity
step one of depth: solving the sterocorrespondance problem
- matching the R and L eye in retina
- tricky because you are trying to match everything (can match random stimuli)
auto-stereogram
- illusions caused when we solve stero-correspondance problem incorrectly (3d image will appear )
- you match the R and L image wrong and then there is incorrect calucation of depth
stage 2 of cacluating depth: calculating the retinal disparity
- once you find the matching retinal images from R and L eye you need to see how different these images are in the L and R eye = retinal/binocular disparity
Retinal/binocular disparity
how different is the object position on the left and right eye
horopter
fixation plane - where it is you are looking
3 possibe outcomes of the retinal dispairty
1) zero disparty
2) crossed disparity
3) uncrossed disparity
zero disparity
- thereis no different in the point in the 2 eyes (light is on the same place in retin)
crossede dispairty
- object is closer to person and in front of the horpoter and light will fall on opposite parts of the eyball and reference point
uncrossed dispairty
- object is behind the horpoter and the light image is not on the same point in both eyes but it is on the same point of the reference point
Pannum’s fusion area
- area that is the limited range for binocular cues of you will see double images if it is too close or too far from heropter
- you an fuse the image from L and R eye to see 3D (when you thumb close to you eye you will see doubles)
Tilt after effect observation
the tilt effect still works even if you adapt one eye and close the other - tilt after effect crosses over to the unadapted eye
- happens because of bincoular cells in the straite cortex (take infro from both eyes = doesn’t matter what eye you use becuase it is the same cell) - cells in straite that are specailized for bincoluar depth
disaprtiy sensitive cells in the visual cortex
- cells can respond to certain disparities (movement in either direction e.g. 0 disparity, looking for specific difference from L and R eyes position)
Strabismus
- related to the msucles that control ocular movement; kids are born with imblance in eye
- can’t do steroimage matching so brain can’t deal with 2 diif. images so suppress one
- if not solved by 6 years old will not develop depth percpetion
Critical period
if no experiences = no development of bincoular cells (only have monocular cells)
Amblyopia
lazy eye - ignoring inform from one eye causes other to not track information anymore (sometimes caused because 1 eye is far/near signed so brain is having trouble bringing 2 eyes to coordination (no depth perception)
- no depth perception if not fixed by 6 years
Bincoluar rivlary
2 diff. images
Chesire cat illusion
- left eye is looking at mirror and R eye look at friend - suddenly you see your friend; brain supresses image of white wall (2 images = brain picks interesting one)
- if it see motion then it is a high priorty stimulus so fingers get molded into your friends face
size constancy problem
- diff. actual sizes give rise to same size on retinal image so actual size can produce differnt retinal sizes (diff. visual angels)
- need to account all visual cues (e.g. depth perception) to achieve size constancy
- you can see things as the same size even the the retinal angle changes
Ponzo illusion
- 2 block on road are same but one looks thicker because depth perceiotn corrects for it e.g. monster
Moon illusion
moon over horizon looks larger than when you are just looking at it in the sky
- because in sky = no depth cues but there are many depth cues on horizon
After image (grid experiment)
- afterimage far away will have a diff. size
Ames room
- using only monocular cues to look at a room so linear perspective is misleading and person walking from one room to the other seems like they are shrinking
Shape constancy
- depending on position of object, the shape of retinal image will vary e.g circle changes into an oval
- differnt shapes produce the same retinal projectional but wee see it as constant shape
Dime box illusion
- dime doesn’t fit in 2d box but in our head we think it does because it seems like 3d
Pointillist reptition
-only know about indiviudal misconnected elements of an object because cones are operating independingly
Goals of object percpetion (2)
1) store in long term memory after perception
2) see something later and see if it matches LTM = recognition
Biederman machine
- only see something for brief # of time and you will recognize it
Top down theory
- starts with knowledge and expectations (can bais how you see ambigous things)
perceptual set
- you can be prepared to see something (memory = expectation drive what you see); but disadvantage when you need to see things that you are not prepared to see
Bottom up theory
- starting from sensory inormation
2 types of processing
1) sptaially parellel
2) serial processing
spatially parellel
- it is everywhere in space at same time
pre attentive processing
- processing doesn’t reuire attention because they respond to all areas without focusing
serial processing
you only get info from one area of visual world at a time (spotllight can only be in one place)
attentional spotlight or focus
- attending to something = ignoring something else
- there is a diff. between looking at something and paying attnetion to it
3 stage model of vision
Retina –> low level vision (botom up level; pre attentive analysis; sptially parellel) –> visual routines (analyses requiring spatial attention; sapital serial) –> visual cognition (object recognition; top down level)
what happens if target and distractor differ by feature
the time it takes to find the target is the same no matter how many distractors there are
- bc of spatial parelle processing (mental process opeates over entire visual field at the same time)
- in low level visual task the more distractors don’t affect time to find object
Gestalt grouping principal
- whole is the difference than the sum of the parts (when we see disconnected objects we want to put them together)
good continuation
idea that we are baised to see continuous lines that go in same direction e.g. 2 lines that cross over vs. 4 separate lines
closure
image looks like a diamond in between 2 walls but could be 2 k’s (we assume things touching are the same object)
formation of illusory contours
when visual system makes good continuation and closure in the Kanizsa trangle/letter stop with white traingle across from it
Figure ground grouping
most famous is Ruben vase vs. face shape - depends on how you group where each line belongs to (what you see as main part and background)
Figure consequences
- is what you remember in your LTM
- smaller things are figure
larger thinkgs will become the background - high spatial frequency = more likely to be seen as figure
object file
a checklist of characteristics e.g. object #1 = black, vertical etc.
combination or conjunction of features
RT on y axis will increase if there are more combination of features beacuse the brain needs to make spatial relatios e.g. what is left and right/inside and outside e.g. the face because it takes up time and attentional spotlight to do visual searchs of each area
Illusory conjunction
when you combine 2 attributes from diff. objects into 1
- occurs with imporper time to see something e.g. colors can spill over to the wrong letter
eye witness testimony
- happens with gun focus and everyone pays attenetion to gun and attributes bystanders to the bank robber
Integrative agnosia
can’t integrate features to make proper object profile e.g. mouth isn’t on the head of the horse bc sptial orientation are not right
Simultagnosia (Balint syndrome)
- can’t differentiate objects from one another inthe same scence
Grouping by common fate
if things move together we form things into an object
biological motion
- from johasson figures you know it is a human
optic flow
movement lines that are streaming across visual field as you go fast (important for balance)
visual capture
you have info from vision but also from vestibular system when vision contradicts vestibular vision is highest prioerty e.g. swinging room experiment
2 systems of vision `
1) image retina system
2) eye head system
image retina system
requires that object changes position from retina e.g. object move from p;lace to place e.g. waterfall illlusion keep eyes steady and will see a after effect
eye head system
- involves eye and head movement - you keep your fovea o goal ball as it moves across the floor so position on retina doesn’t change
ambert fleischl effect
- track image with eyes or keep them still - keeping still seems like pendaulum is going faster because tracking object tracks the pace your eyeballs are moving