Cynthia Breazeal Aaron Edsinger Paul Fitzpatrick Brian Scassellati MIT AI Lab Social Constraints on...
-
Upload
archibald-tate -
Category
Documents
-
view
215 -
download
0
Transcript of Cynthia Breazeal Aaron Edsinger Paul Fitzpatrick Brian Scassellati MIT AI Lab Social Constraints on...
![Page 1: Cynthia Breazeal Aaron Edsinger Paul Fitzpatrick Brian Scassellati MIT AI Lab Social Constraints on Animate Vision.](https://reader036.fdocuments.in/reader036/viewer/2022062802/56649ea05503460f94ba35ba/html5/thumbnails/1.jpg)
Cynthia BreazealCynthia Breazeal
Aaron EdsingerAaron Edsinger
Paul FitzpatrickPaul Fitzpatrick
Brian ScassellatiBrian Scassellati
MIT AI LabMIT AI Lab
Social Constraints Social Constraints on on
Animate VisionAnimate Vision
![Page 2: Cynthia Breazeal Aaron Edsinger Paul Fitzpatrick Brian Scassellati MIT AI Lab Social Constraints on Animate Vision.](https://reader036.fdocuments.in/reader036/viewer/2022062802/56649ea05503460f94ba35ba/html5/thumbnails/2.jpg)
2Humanoids2000
Social constraints
Robots create expectations through their physical form – particularly humanoid robots
But with careful use, these expectations can facilitate smooth, intuitive interaction– Provide a natural “vocabulary” to make the robot’s
behavior and state readable by a human– Provide natural frameworks for trying to negotiate a
change in each other’s behavior and (through readability) knowing when you have succeeded
– These elements have their own internal logic and constraints which, if violated, lead to confusion
![Page 3: Cynthia Breazeal Aaron Edsinger Paul Fitzpatrick Brian Scassellati MIT AI Lab Social Constraints on Animate Vision.](https://reader036.fdocuments.in/reader036/viewer/2022062802/56649ea05503460f94ba35ba/html5/thumbnails/3.jpg)
3Humanoids2000
Visually-mediated social elements
Readable locus of attention Negotiation of the locus of attention Readable degree of engagement Negotiation of interpersonal distance Negotiation of object showing Negotiation of turn-taking timing
Readable locus of attention
![Page 4: Cynthia Breazeal Aaron Edsinger Paul Fitzpatrick Brian Scassellati MIT AI Lab Social Constraints on Animate Vision.](https://reader036.fdocuments.in/reader036/viewer/2022062802/56649ea05503460f94ba35ba/html5/thumbnails/4.jpg)
4Humanoids2000
Readable locus of attention
Attention can be deduced from behavior Or can be expressed more directly
![Page 5: Cynthia Breazeal Aaron Edsinger Paul Fitzpatrick Brian Scassellati MIT AI Lab Social Constraints on Animate Vision.](https://reader036.fdocuments.in/reader036/viewer/2022062802/56649ea05503460f94ba35ba/html5/thumbnails/5.jpg)
5Humanoids2000
Kismet – a readable robot
Eye tilt
Left eye panRight eye pan
Camera with wide field of view
Camera with narrow field of view
Designed to evoke infant-level social interactions
– Eibl-Eiblsfeldt “baby scheme”– physical size, stature
But not exactly human infant– caricature that is readable
Naturally elicit scaffolding acts characteristic of parent-infant scenarios
– directing attention– affective feedback, reinforcement– simplified behavior, suggested to
make perceptual task easier– slow down, go at infant’s pace
![Page 6: Cynthia Breazeal Aaron Edsinger Paul Fitzpatrick Brian Scassellati MIT AI Lab Social Constraints on Animate Vision.](https://reader036.fdocuments.in/reader036/viewer/2022062802/56649ea05503460f94ba35ba/html5/thumbnails/6.jpg)
6Humanoids2000
Visually-mediated social elements
Readable locus of attention Negotiation of the locus of attention Readable degree of engagement Negotiation of interpersonal distance Negotiation of object showing Negotiation of turn-taking timing
![Page 7: Cynthia Breazeal Aaron Edsinger Paul Fitzpatrick Brian Scassellati MIT AI Lab Social Constraints on Animate Vision.](https://reader036.fdocuments.in/reader036/viewer/2022062802/56649ea05503460f94ba35ba/html5/thumbnails/7.jpg)
7Humanoids2000
Negotiating the locus of attention
For object-centered activities, attention is fundamental There are natural strategies people use to direct attention The robot’s attention must be receptive to these influences,
but also serve the robot’s own agenda
One person’s strategies Another’s strategies
![Page 8: Cynthia Breazeal Aaron Edsinger Paul Fitzpatrick Brian Scassellati MIT AI Lab Social Constraints on Animate Vision.](https://reader036.fdocuments.in/reader036/viewer/2022062802/56649ea05503460f94ba35ba/html5/thumbnails/8.jpg)
8Humanoids2000
Skin tone Color Motion Habituation
Weightedby behavioral
relevance
Pre-attentive filters
External influences on attention
Attention is allocated according to salience Salience can be manipulated by shaking an object,
bringing it closer, moving it in front of the robot’s current locus of attention, object choice, hiding distractors, …
Current input Saliency map
![Page 9: Cynthia Breazeal Aaron Edsinger Paul Fitzpatrick Brian Scassellati MIT AI Lab Social Constraints on Animate Vision.](https://reader036.fdocuments.in/reader036/viewer/2022062802/56649ea05503460f94ba35ba/html5/thumbnails/9.jpg)
9Humanoids2000
Tuned to natural cues
stimulus category
stimulus presentationsaverage time (s)
commonly used cues
commonly read cues
color and movement
yellow dinosaur 8 8.5
motion across
centerline,
shaking,
bringing object close
change in visual
behavior,
face reaction,
body posture
multi-colored block
8 6.5
green cylinder 8 6.0
movement only
black&white cow
8 5.0
skin toned and movement
pink cup 8 6.5
hand 8 5.0
face 8 3.0
Overall 56 5.8
![Page 10: Cynthia Breazeal Aaron Edsinger Paul Fitzpatrick Brian Scassellati MIT AI Lab Social Constraints on Animate Vision.](https://reader036.fdocuments.in/reader036/viewer/2022062802/56649ea05503460f94ba35ba/html5/thumbnails/10.jpg)
10Humanoids2000
Can shape an interaction
The robot’s attention can be manipulated repeatedly So caregiver can shape an interaction into the form of an
object-centered game, or a teaching session
![Page 11: Cynthia Breazeal Aaron Edsinger Paul Fitzpatrick Brian Scassellati MIT AI Lab Social Constraints on Animate Vision.](https://reader036.fdocuments.in/reader036/viewer/2022062802/56649ea05503460f94ba35ba/html5/thumbnails/11.jpg)
11Humanoids2000
“Seek face” –high skin gain, low color saliency gainLooking time 28% face, 72% block
“Seek toy” –low skin gain, high saturated-color gain
Looking time 28% face, 72% block
Internal influences on attention
Internal influences bias how salience is measured The robot is not a slave to its environment
![Page 12: Cynthia Breazeal Aaron Edsinger Paul Fitzpatrick Brian Scassellati MIT AI Lab Social Constraints on Animate Vision.](https://reader036.fdocuments.in/reader036/viewer/2022062802/56649ea05503460f94ba35ba/html5/thumbnails/12.jpg)
12Humanoids2000
Maintaining visual attention
Want attention to be persistent enough to permit coherent behavior
Must be able to maintain fixation on an object, when behaviorally appropriate
Attention system interacts closely with tracker to support this robustly
![Page 13: Cynthia Breazeal Aaron Edsinger Paul Fitzpatrick Brian Scassellati MIT AI Lab Social Constraints on Animate Vision.](https://reader036.fdocuments.in/reader036/viewer/2022062802/56649ea05503460f94ba35ba/html5/thumbnails/13.jpg)
13Humanoids2000
Visually-mediated social elements
Readable locus of attention Negotiation of the locus of attention Readable degree of engagement Negotiation of interpersonal distance Negotiation of object showing Negotiation of turn-taking timing
![Page 14: Cynthia Breazeal Aaron Edsinger Paul Fitzpatrick Brian Scassellati MIT AI Lab Social Constraints on Animate Vision.](https://reader036.fdocuments.in/reader036/viewer/2022062802/56649ea05503460f94ba35ba/html5/thumbnails/14.jpg)
14Humanoids2000
Readable degree of engagement
Visual behavior conveys degree of commitment– fleeting glances– smooth pursuit– full body orientation
Gaze direction, facial expression, and body posture convey robot’s interest
![Page 15: Cynthia Breazeal Aaron Edsinger Paul Fitzpatrick Brian Scassellati MIT AI Lab Social Constraints on Animate Vision.](https://reader036.fdocuments.in/reader036/viewer/2022062802/56649ea05503460f94ba35ba/html5/thumbnails/15.jpg)
15Humanoids2000
Visually-mediated social elements
Readable locus of attention Negotiation of the locus of attention Readable degree of engagement Negotiation of interpersonal distance Negotiation of object showing Negotiation of turn-taking timing
![Page 16: Cynthia Breazeal Aaron Edsinger Paul Fitzpatrick Brian Scassellati MIT AI Lab Social Constraints on Animate Vision.](https://reader036.fdocuments.in/reader036/viewer/2022062802/56649ea05503460f94ba35ba/html5/thumbnails/16.jpg)
16Humanoids2000
Comfortable interaction distance
Too close – withdrawal response
Too far – calling
behavior
Person draws closer
Person backs off
Beyond sensor range
Negotiating interpersonal distance
Robot establishes a “personal space” through expressive cues
Tunes interaction to suit its vision capabilities
![Page 17: Cynthia Breazeal Aaron Edsinger Paul Fitzpatrick Brian Scassellati MIT AI Lab Social Constraints on Animate Vision.](https://reader036.fdocuments.in/reader036/viewer/2022062802/56649ea05503460f94ba35ba/html5/thumbnails/17.jpg)
17Humanoids2000
Negotiating interpersonal distance
Robot backs away if person comes too close Cues person to back away too – social amplification Robot makes itself salient to call a person closer if too far away
“Back off buster!” “Come hither, friend”
![Page 18: Cynthia Breazeal Aaron Edsinger Paul Fitzpatrick Brian Scassellati MIT AI Lab Social Constraints on Animate Vision.](https://reader036.fdocuments.in/reader036/viewer/2022062802/56649ea05503460f94ba35ba/html5/thumbnails/18.jpg)
18Humanoids2000
Visually-mediated social elements
Readable locus of attention Negotiation of the locus of attention Readable degree of engagement Negotiation of interpersonal distance Negotiation of object showing Negotiation of turn-taking timing
![Page 19: Cynthia Breazeal Aaron Edsinger Paul Fitzpatrick Brian Scassellati MIT AI Lab Social Constraints on Animate Vision.](https://reader036.fdocuments.in/reader036/viewer/2022062802/56649ea05503460f94ba35ba/html5/thumbnails/19.jpg)
19Humanoids2000
Negotiating object showing
Robot conveys preferences about how objects are presented to it through irritation, threat responses
Again, tunes interaction to suit its limited vision Also serves protective role
Comfortable interaction speed
Too fast – irritation response
Too fast,Too close –
threat response
![Page 20: Cynthia Breazeal Aaron Edsinger Paul Fitzpatrick Brian Scassellati MIT AI Lab Social Constraints on Animate Vision.](https://reader036.fdocuments.in/reader036/viewer/2022062802/56649ea05503460f94ba35ba/html5/thumbnails/20.jpg)
20Humanoids2000
Negotiating object showing
Robot “shuts out” close, fast moving object – threat response Robot backs away if object too close Robot cranes forward as expression of interest
Threat response Withdrawal, startle
![Page 21: Cynthia Breazeal Aaron Edsinger Paul Fitzpatrick Brian Scassellati MIT AI Lab Social Constraints on Animate Vision.](https://reader036.fdocuments.in/reader036/viewer/2022062802/56649ea05503460f94ba35ba/html5/thumbnails/21.jpg)
21Humanoids2000
Visually-mediated social elements
Readable locus of attention Negotiation of the locus of attention Readable degree of engagement Negotiation of interpersonal distance Negotiation of object showing Negotiation of turn-taking timing
![Page 22: Cynthia Breazeal Aaron Edsinger Paul Fitzpatrick Brian Scassellati MIT AI Lab Social Constraints on Animate Vision.](https://reader036.fdocuments.in/reader036/viewer/2022062802/56649ea05503460f94ba35ba/html5/thumbnails/22.jpg)
22Humanoids2000
Turn-Taking
Cornerstone of human-style communication, learning, and instruction
Four phases of turn cycle– relinquish floor
– listen to speaker
– reacquire floor
– speak
Integrates– visual behavior & attention
– facial expression & animation
– body posture
– vocalization & lip synchronization
![Page 23: Cynthia Breazeal Aaron Edsinger Paul Fitzpatrick Brian Scassellati MIT AI Lab Social Constraints on Animate Vision.](https://reader036.fdocuments.in/reader036/viewer/2022062802/56649ea05503460f94ba35ba/html5/thumbnails/23.jpg)
23Humanoids2000
Examples of turn-taking
Turn-taking is fine grained regulation of human’s behavior Uses envelope displays, facial expressions, shifts of gaze and body posture Tightly coupled dynamic of contingent responses to other
Kismet and Adrian Kismet and Rick
![Page 24: Cynthia Breazeal Aaron Edsinger Paul Fitzpatrick Brian Scassellati MIT AI Lab Social Constraints on Animate Vision.](https://reader036.fdocuments.in/reader036/viewer/2022062802/56649ea05503460f94ba35ba/html5/thumbnails/24.jpg)
24Humanoids2000
Evaluation of Performance
Naive subjects– ranging in age from 25 to 28
– All young professionals.
– No prior experience with Kismet
– video recorded
Turn-taking performance– 82% “clean” turn transitions
– 11% interruptions
– 7% delays followed by prompting
Significant flow disturbances– tend to occur in clusters
– 6.5% of the time, but rate diminishes
Evidence for entrainment– shorter phrases
– wait longer for response
– read turn-taking cues
– 0.5—1.5 seconds between turns
37+8:06 – 8:43end @ 8:43
447:18 – 8:02
216:54 – 7:15
76:43 – 6:50start @ 6:43subject 2
37+17:30 – 18:07end @ 18:07
7016:20 – 17:25
1915:56 – 16:15
2115:37 – 15:54
1315:20 – 15:33start @ 15:20subject 1
subject 3
80+9:20 – 10:40end @ 10:40 min
458:25 – 9:10
587:18 – 8:16
186:58 – 7:16
536:00 – 6:53
245:30 – 5:54
155:08 – 5:23
104:52 – 4:58start @ 4:52 min
seconds between
disturbances
time stamp (min:sec)
![Page 25: Cynthia Breazeal Aaron Edsinger Paul Fitzpatrick Brian Scassellati MIT AI Lab Social Constraints on Animate Vision.](https://reader036.fdocuments.in/reader036/viewer/2022062802/56649ea05503460f94ba35ba/html5/thumbnails/25.jpg)
25Humanoids2000
Conclusion
Active vision involves choosing a robot’s pose to facilitate visual perception.
Focus has been on immediate physical consequences of pose.
For anthropomorphic head, active vision strategies can be “read” by a human, assigned an intent which may then be completed beyond the robot’s immediate physical capabilities.
Robot’s actions have communicative value, to which human responds.
25Humanoids2000