Lecture 3 - Announcement Foundations of Mind ! 200.336 !...

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: Announcement Foundations of Mind ! 200.336 ! "There is a student in this class who requires the services of a note taker. " This is a paid position through the Student Disability Services Office. " " If you are interested in performing this service, please send an e-mail to Notetaking@jhu.edu" " The note taking accommodation is to supplement the personal notes of the registered student. It is not intended to substitute the student's own notes, attendance, or class participation.”" ! ! 2/7/2012 ! Lecture #3 ! ! Depth Perception 2 ! 2 ! Signatures of Depth Percep5on in Infants? Signatures of Depth Percep5on in Infants? 1 addi5onal control: Repeat the stereo preference experiment without the glasses (both images seen by both eyes- - no stereo) Stereopsis: conclusions vs. Real stereo Double images No preference for double images in the no- stereo condi5on. Argues against Berkeley “simply confused” worry 3 3- 4 month old infants show all the key psychophysical signatures of stereopsis in adults. The signatures emerge abruptly. The signatures of stereopsis emerge before reaching, sugges5ng that they aren’t learned by associa5on with touch (a la Berkeley) (though visual experience may influence their development). Humans may be predisposed to perceive depth from binocular disparity. (a la Descartes) 4 1 Two Approaches to Solving the Problem Mul5ple Cues to Depth 1) Psychophysical signatures of depth percep5on in infants 2) Equivalence of different cues to depth Linear perspective Interposition (occlusion) 5 Mul5ple Cues to Depth 6 ! The Equivalence of Depth Cues We must learn to associate the sensa5ons of each of these cues to the tac5le experience of depth! The mind automa5cally treats all of these as signifying info about the same thing in the world: Depth! Optical expansion (a motion cue) 7 ! 8 2 The Equivalence of Depth Cues The Equivalence of Depth Cues The same depth arrangement can be specified by different cues (Yonas studies) Mo5on (accre5on/dele5on) The same depth arrangement can be specified by different cues (Yonas studies) Stereo Mo5on 9 The Equivalence of Depth Cues Method #1: Looking Time Problem: How can we test if infants see these as the “same”?? 10 Developmental Method #3: Habitua5on Method: Visual habitua5on procedure Seconds Method #2: Preferen5al Looking Stereo (accre5on/dele5on) How should young infants perceive the two arrangements? Perceive as Perceive as the SAME! DIFFERENT! Method #3: Habitua5on For example… 11 12 3 Developmental Method #3: Habitua5on Developmental Method #3: Habitua5on Seconds Method: Visual habitua5on procedure Seconds Method: Visual habitua5on procedure 13 Developmental Method #3: Habitua5on 14 Developmental Method #3: Habitua5on Method: Visual habitua5on procedure Method: Visual habitua5on procedure Ho hum… Habitua5on curve Seconds Seconds Habitua5on curve 15 16 4 Developmental Method #3: Habitua5on Developmental Method #3: Habitua5on Method: Visual habitua5on procedure Habitua5on display: depth from MOTION (4- month infants) Test displays: depth from STEREO (triangle in front for half the babies, behind for the rest) Wow!!! and Dishabituation! looking time (sec) Seconds Habitua5on curve Continued! habituation! 17 50 45 40 35 30 25 20 15 10 5 0 Novel Depth Familiar Depth 1 2 3 Habitua5on 4 5 6 Test 18 The Equivalence of Depth Cues 1) Psychophysical signatures of depth percep5on in infants 2) Equivalence of different cues to depth •  These two approaches suggest that infants see in depth by 3 mos •  Provides some support for Descartes, and innateness of depth percep5on •  3- 4- month olds =youngest babies we can test using looking methods •  3- 4 months of experience is, perhaps, quite a lot….? •  Ques5on: Are there other methods that may reveal earlier depth percep5on in younger infants or other animals? 19 Foundations Of Mind Today: Visual Cliff Next Time: Size and Shape Constancy Slide# 20 ! 5 Next approach: Eleanor J. Gibson and the comparative study of perceptual development (1910-2002) Assumption: many basic perceptual capacities evolved before humans and so are shared by humans and other animals. Strategy: (1) study space perception in a wide range of animals. If the same capacities appear across different species, the capacities probably are shared by humans. (2) study space perception in precocial animals (e.g., those who locomote at birth) and in animals reared under controlled conditions. If space perception is innate, it should appear at birth and without visual Slide# 21 ! experience. Slide# 23 ! The visual cliff (Gibson & Walk, 1958) Gibson’s incidental discovery: newborn goats on a pedestal. platform glass vision or touch? texture Slide# 22 ! Slide# 24 ! 6 per cent crossing Is this effect specific to animals that live on mountains? 100 90 80 70 60 50 40 30 20 10 0 Shallow side Deep side All goats crossed the shallow side. No goats crossed the deep side. shallow deep The Problem of Depth Strategy: Test for innateness by using animal models Precocial animals No way! baby goats An innate capacity to perceive and avoid visual drop-offs. 25! Slide# Slide# 26 ! The visual cliff: Further studies (1) Is this effect specific to animals that live on mountains? Tests of other animals: lambs, chickens, rats, turtles, cats, dogs, humans. Findings: Cliff avoidance in every animal tested. Cliff avoidance appears as soon as an animal is capable of visually guided locomotion: goats, chicks, lambs: at birth rats and cats: weeks after birth humans??? Slide# 27 ! Reflexes Reflex: Stepping Stimulation: Supported infant’s bare feet touch surface Response: Alternating stepping response Probable function: Preparation for voluntary walking Slide# 28 ! 7 Motor Milestones: Gross Development Motor Milestones: Gross Development Slide# 29 ! Motor Milestones: Gross Development Slide# 30 ! A Continuity Assumption Other things being equal, basic mechanisms and capacities will be preserved over the course of evolution and therefore will be common to humans and other animals. true: heart, lungs, eyes, ears, breathing, … false: bi-pedal motion, talking, ... Is it true of cliff avoidance? Cephalocaudal trend: motor control proceeds from head downwards Slide# 31 ! Slide# 32 ! 8 The visual cliff in human infants per cent crossing 9-month old Infants crossing the visual cliff 100 90 80 70 60 50 40 30 20 10 0 Slide# 33 ! The Problem of Depth shallow deep infants Slide# 34 ! The Problem of Depth So which depth cues do babies/animals use?? Question: Can infants use any of these cues to perceive depth??? Yikes! 3-month olds Campos: When pre-crawlers Are lowered onto deep side, Heart rate changes… Infants perceived depth even prior to crawling experience! Possible depth cues: •  Ocular cues (convergence) • Motion cues (motion parallax) • Monocular (e.g., texture) • Binocular (e.g., disparity) Amount of retinal motion from sideways head movement Deep side Shallow side & center board Slide# 35 ! Slide# 36 ! 9 The Problem of Depth The Problem of Depth So which depth cues do babies/animals use?? Possible depth cues: •  Ocular cues (convergence) • Motion cues (motion parallax) •  Pictorial cues (texture size smaller on deep side) So which depth cues do babies/animals use?? Amount of retinal motion from sideways head movement Deep side Experimental tests: 1) Monocular viewing does not impair cliff avoidance: Binocular cues not necessary 2) Texture size does not affect cliff avoidance at birth: Texture differences not necessary Retinal image of shallow side 3) Removing texture altogether/prohibiting movement eliminates cliff avoidance: Motion cues provide the critical information Retinal image of deep side Shallow side & center board Slide# 37 ! The visual cliff: Controlled rearing studies (2) Do animals (rats, cats, humans) learn to perceive the visual spatial layout prior to locomotion? Further studies of rats and cats: --Newborn animals randomly assigned to two groups. --The two groups are treated identically except for visual experience: light-reared vs. dark-reared. --For testing, all animals introduced into the light and placed immediately on the cliff. Slide# 39 ! Slide# 38 ! The Problem of Depth Strategy: Test for innateness by using animal models Controlled rearing conditions Light-reared (normal) rats Dark-reared rats Slide# 40 ! 10 The Problem of Depth per cent crossing Rats crossing the deep side of the cliff Slide# 41 ! The visual cliff: Controlled rearing studies of cats per cent willing Cats willing to sit on the deep side A cat sitting on the deep side 100 90 80 70 60 50 40 30 20 10 0 deep side day 1 light- reared day 1 dark- reared day 3 dark- reared •  Dark-reared, reintroduced to light at age 3-4 weeks •  Immediately after dark-rearing, no cliff avoidance. •  After 2-3 days in light, normal avoidance. Slide# 43 ! 100 90 80 70 60 50 40 30 20 10 0 shallow side deep side light-reared dark-reared • Cliff avoidance in both dark-reared and light-reared animals • No reliable difference between the two groups Slide# • Cliff avoidance is independent of visual experience in rats 42! Experience effects in cats: What happens in the first 3 days in the light? Two possible answers: --Experience teaches the cat that visual cliffs are dangerous. Specific effect of experience. --Experience tunes up the cat’s visual system and allows innate capacities to perceive depth and avoid dropoffs to be expressed. Non-specific effect of experience. Gibson’s test: controlled rearing on the deep side of the cliff. --if specific effect, cat will learn that the cliff is safe. --if non-specific effect, cat will begin to avoid the cliff. Finding: normal cliff avoidance after 3 days. Cliff avoidance depends on non-specific visual experience but not on learning that visible dropoffs are dangerous. Slide# 44 ! 11 More experience effects in cats: Held & Hein’s studies of active vs. passive locomotion. per cent crossing Active- and passive-reared cats crossing the visual cliff 100 90 80 70 60 50 40 30 20 10 0 shallow side deep side active-reared passivereared Cliff avoidance depends on experience with active locomotion. Slide# 45 ! Compare: prism adaptation in human adults. The visual cliff: Conclusions Across a range of species, cliff avoidance emerges at the onset of locomotion Cliff avoidance is independent of any visual experience in some animals (goats, chicks, rats) Cliff avoidance depends on active locomotion and non-specific visual experience in other animals, but is independent of specific experience with cliffs (cats) Cliff avoidance appears in humans as soon as a baby crawls and can be observed even before then by measuring heart rate Slide# 46 ! The Problem of Depth The Problem of Depth So how do we ever figure out how far away objects are? So how do we ever figure out how far away objects are? Cue # 2: Binocular disparity The closer an object is, the more different the image received by the left vs. right eye Cue # 2: Binocular disparity The closer an object is, the more different the image received by the left vs. right eye Binocular cue Slide# 47 ! Binocular cue Slide# 48 ! 12 The Problem of Depth The Problem of Depth So how do we ever figure out how far away objects are? So how do we ever figure out how far away objects are? Cue # 3: Cue # 4: Optical expansion: A sudden increase in apparent object size may signal that the object is approaching Motion parallax: Nearby objects appear to “pass by” at a faster rate than more distant objects Motion cue Motion cue Slide# 49 ! Slide# 50 ! The Problem of Depth The Problem of Depth So how do we ever figure out how far away objects are? So how do we ever figure out how far away objects are? Cue # 4: Cue # 5: Motion parallax: Nearby objects appear to “pass by” at a faster rate than more distant objects Interposition: Nearby objects are usually more complete; farther away objects are often occluded Motion cue Monocular/ pictoral cue Slide# 51 ! Slide# 52 ! 13 The Problem of Depth The Problem of Depth So how do we ever figure out how far away objects are? So how do we ever figure out how far away objects are? Cue # 6: Cue # 7: Texture: Objects that are farther form a finer texture than nearby objects Linear perspective: Lines appear to converge with increasing distance Monocular/ pictoral cue Monocular/ pictoral cue Slide# 53 ! Slide# 54 ! 14 ...
View Full Document

This note was uploaded on 02/26/2012 for the course PSYCHOLOGY 336 taught by Professor Halberda during the Spring '11 term at Johns Hopkins.

Ask a homework question - tutors are online