Presentation is loading. Please wait.

Presentation is loading. Please wait.

By Mari Riess Jones The Dynamics of Attending and Scene Analysis.

Similar presentations


Presentation on theme: "By Mari Riess Jones The Dynamics of Attending and Scene Analysis."— Presentation transcript:

1

2 By Mari Riess Jones The Dynamics of Attending and Scene Analysis

3 Thanks to Ralph Barnes Devin McAuley Heather Moynihan Noah MacKenzie Amandine Penel Jennifer Puente Nate Vaughan Special Thanks to Edward Large Edward Large

4 I.Scene Analysis: A Neo Gestalt view II. Scene Analysis: A dynamic perspective III. The Changing Scene: Tracking in real timeOverview

5 I. Scene Analysis: A Neo Gestalt view

6 What’s a Scene? scene The term scene is taken from vision. Two cases are most common: 1. A stationary viewer 2. A moving viewer The scene is the static visual layout. The scene is the static visual layout.

7 Sound Scenes A common assumption: We use the proximal sound stimulus to recover the location and/or identity of a distal, vibrating, object within a scene. This is the source identity principle.

8 Bregman’s Theory of Scene Analysis Bregman builds on this idea: Goal of auditory perception… is identification of one or more sources of a proximal mixture of sounds. If an unfolding sound pattern is perceived as a structurally intact, one stream i.e. as one stream, ONE source. then it must come from ONE source.

9 Other patterns? streams, If a pattern is perceived to break up, i.e. form N different subjective units, streams, then, it must … come from N separate sources. Patterns NOT perceived as intact ….

10 Time high low Another source? Example: A sound pattern frequency A stream ? One source?

11 The Big Question seem What makes a pattern seem structurally intact? stream? Or… What determines a stream?

12 Bregman’s Answer Involves a two stage theory in which Neo Gestalt principles are critical.

13 The Neo Gestalt View I.Primitive processes Innate, automatic, effortless, Determine coherent perceptual groups at fast rates. II. Schema driven processes Learned, voluntary, effortful, Determine attention to groupings at slow rates. Two Stages

14 Auditory Scene Perception Stage I: Primitive Processes Operate on Gestalt grouping principles: proximity in frequency and time; similarity (e.g., timbre); continuity; exclusive allocation… and so forth These are domain general ( I.e. speech, music, natural environs,etc.)

15 Proximity vs Continuity Proximity vs Continuity Tougas and Bregman (1985) proximitycontinuity. In a series of famous studies pitted proximity against continuity. Goal: To assess which principle dominated, and to dispute a trajectory explanation. Bouncing versus Crossing trajectories

16 Time in seconds Possible Percepts

17 Conclusion Proximity in frequency dominates People hear bouncing patterns

18 Rhythm Perception interleaved streams Speeding up a sequence can break it, perceptually, into interleaved streams This changes its perceived rhythm! This changes its perceived rhythm! rhythm A stream usually conveys a rhythm

19 Rhythm is perceived to have a ‘galloping ‘rhythm This slow melody… Time relationship Rhythm perception… depends on the relationship of frequency change (  f ) to time change (  t ). Hifrequency Lo frequency Lo frequency ff t t 2t2t

20 But… That Rhythm Disappears In a faster melody with the same tones. isochronous rhythm. Now we hear two melodies (streams) each with an isochronous rhythm.

21 A Neo Gestalt Explanation Builds on Gestalt proximity: 1.Temporal proximity 1.Temporal proximity is heightened at fast rates. frequency proximity 2.But frequency proximity dominates. 3.Two streams 3.Two streams form based on pitch proximity: A high tone stream and a low tone stream.

22 II.Scene Analysis: A Dynamic Perspective

23 Auditory Patterns Most of our environment comprises sources that turn on and off and modulate over time to project pattern structure that is…. motion-like and rhythmic

24 Important Pattern Properties 1.Motionlike properties: Rates of change. [ e.g.,  f/  t ] 2. Rhythmic properties: Relative timing. [e.g.,  ti/  tk ]

25 Motionlike Trajectories Pitch Velocity Pitch Velocity:  f /  t locity. In 1976, I proposed that the rate-of-change of frequency (log scale) may be perceived as a sort of velocity.

26 Time high low Determinants of Pitch Velocity Determinants of Pitch Velocity frequency A sudden increase in rate-of-change in an isochronous rhythm is perturbing. ff tt  f/  t

27 Motionlike Trajectory limits on pitch velocity?  f log FAST  t Slow Intact Serial Percepts Velocity Limit ( Jones, 1976) RATE SMALL LARGE Stimulus patterns streamed

28 Rhythmic Properties: Relative Timing  f log FAST  t =IOI SLOW Limit RATE Slower patterns

29 Alternative View: A focus on time How do rhythmic properties of a pattern affect real time responding to it? In the remainder of this talk, I address this question.

30 General Assumptions Connecting through time Pattern timing determines how living things connect with a changing scene. Pattern timing determines how living things connect with a changing scene. Living things Animate creatures possess internal periodicities that can synchronize to patterns. Animate creatures possess internal periodicities that can synchronize to patterns. E nvironment Scenes often project dynamic patterns with relative timing that is quasi- periodic.

31 Great Gray Owl Horned Lark Cardinal  t = IOI = 700 ms Some quasi-periodic patterns

32 Internal Periodicities: Biological Oscillations Most studied are circadian rhythms; oscillatory periods of ca. 24 hours. Most studied are circadian rhythms; oscillatory periods of ca. 24 hours. Many internal periodicities exist; periods of 1 ms, several minutes, years !. Many internal periodicities exist; periods of 1 ms, several minutes, years !.

33 Connecting “How do internal oscillations of an organism connect with pattern timing? “ One answer: By synchronizing. More precisely: By entrainment More precisely: By entrainment

34 Biological Entrainment Pittendrigh & Daan (1976). Circadian rhythms of 24 hours synchronize to light/dark cycles. Pittendrigh & Daan (1976). Circadian rhythms of 24 hours synchronize to light/dark cycles. Daan et al. (2001) recently updated this two oscillator model. Daan et al. (2001) recently updated this two oscillator model.

35 Summer Winter Winter From Daan et al. J.Biological Rhythms (2001) Example: Two Entraining Oscillations M Phase locked E Phase locked

36 Shorter Periods? In species with refined use of fast sound patterns, Can we postulate, internal oscillations with periods less than the circadian period ?

37 Generalizing….. Long natural periods Long natural periods Respond to light changes Respond to light changes Most Biological Oscillators Oscillators Alternatively Alternatively Other Oscillators Other Oscillators Short natural periods Respond to sound changes changes

38 Background: Rhythmic (periodic) Expectancies Quasi-periodicities in Quasi-periodicities in Rhythmical sound patterns Rhythmical sound patterns Induce Periodic expectancies Induce Periodic expectancies Quasi-periodicities in Quasi-periodicities in Rhythmical sound patterns Rhythmical sound patterns Induce Periodic expectancies Induce Periodic expectancies Perhaps such expectancies reflect entrainments to patterns of stimulus IOIs?

39 Back to Synchrony: A simple premise When we attend, an internal activity is elicited … that co-occurs with the attended object. When we attend, an internal activity is elicited … that co-occurs with the attended object. This activity involves Internal oscillations (Jones, 1976). This activity involves Internal oscillations (Jones, 1976).

40 Where elements appear… then disappear Attending to Dynamic Patterns Time Attention must occur in synchrony with elements.

41 Dynamic Attending Selective attending in time is paced by stimulus timing. Postulates that : Attentional focus in time becomes phase-locked to elements.

42 III. The Changing Scene: Tracking sounds in real time

43 Attending oscillates in time internal oscillations. Resources are periodically distributed as internal oscillations. Time Pulses of attention energy Inter-onset-interval, IOI

44 Expected point in time Phase Period Formalized as an Oscillator Model Formalized as an Oscillator Model : by Ed Large (Large & Jones, 1999) Stimulus timing paces the oscillator. Oscillator period and phase determine this entrainment. Periodic Expectancies Stimulus IOI

45 Phase,  iPeriod, Pi Tones, i Period and Phase change: They draw attending rhythms to “fit” a time pattern. Period changes Phase changes adaptive The adaptive oscillator

46 The entrained oscillator gravitates to an attractor state of: 1. Phase Synchrony: 1. Phase Synchrony: A zero time difference between an sound onset and a pulse peak. 2. Period matching: 2. Period matching: A zero time difference between an IOI and oscillator period (limit cycle). A simple dynamical system.

47 Relative Phase,  i Relative Phase,  i Gives observed - expected times  i  i is the order variable in this system. It summarizes the state of the system at any instant i.  i+1 =  i + IOI i+1 /P i -   F(  i ) The coupling term, F, here is sinusoidal: F(  )= {1/2  }sin 2  => 

48  -  Phase Attractor: Synchrony Relative Phase, i Entrainment Region {   = 1.0 } Theoretical Phase Attractor: synchrony

49 What does this approach purchase ? A few experiments ….

50 Preview: Our Experimental Designs Will involve a context rhythm, given by tones: Followed by a comparison, C, (of some sort) Time Judgment Comparison IOI Pitch Judgment Comparison tone C

51 Time Judgment Tasks Rationale If relative timing in sound arrays entrains attending, then the context rhythm should Shape temporal expectancies Affect time judgments.

52 Example Study: Example Study: Judgment Task Judgment Task: Is a comparison time interval “shorter”, “same” or “ longer” ? “ Ignore context IOIs“ “ Ignore context IOIs“ Context IOI Standard IOI Comparison IOI Context IOIs are 600 ms. This rhythm should synchronize attending. +/-  T

53 Standard Ending Varies: Standard Ending Varies: Expected Standard Ending: = 600 ms Unexpected Standard Endings: Early = 579 ms Late = 621 ms Very Early = 524 ms Very Late = 676 ms 23 A comparison IOI is always yoked to standard IOI

54 STANDARD IOI Expected Ending Expected Ending: “Same” Phase COMPARISON IOI Period Modeling Oscillatory Attending Context IOIs 600 ms 600 When a standard ends as expected, the oscillator’s period is accurately preserved ….as a memory of the standard IOI.

55 The Reason: The Attentional Pulse The Reason: The Attentional Pulse Pulse location depends on rhythmic context Expected time 15 Pulse location

56 One Prediction: Expectancy Profile A quadratic accuracy Expectancy Profile: Attentional Pulse

57 Predicted and Observed Expectancy Profiles (Large & Jones 1999) P <.01

58 Subsequent Experiments Over a dozen experiments later (e.g., Barnes & Jones, 2000; McAuley & Jones, 2002) We Find: The profile vanishes with no context; sharpens with longer sequences. Holds for different rates. Reveals harmonic scaling of timing.

59 Different Induction Rates Three conditions: different induction rates. Condition Induction IOIs Standard durations ms 300 ms524, 600, 676 ms ms 500 ms 524, 600, 676 ms ms 600 ms524, 600, 676 ms Isochronous sequences This task required same vs different duration judgments

60 Observed Expectancy Profiles Three Time Judgment Conditions

61 Conclude Back to Relative timing Ratios induction IOIsexpected standards 1. Ratios of induction IOIs to expected standards seem influential. 2. Harmonic ratios 2. Harmonic ratios of 1:1 and 1:2 yield different profiles than the 500 ms with a more complex ratio.

62 Pitch Judgments We generalized this approach to attending to tone frequency. Does rhythmic context affect pitch judgments?

63 Standard pitch Comparison pitch Task Task: Is a comparison pitch: “same”, “higher”, ”lower” than a standard pitch? “Ignore distracting pitches!” “higher” same” “lower ” A Pitch Judgment Task A Pitch Judgment Task Critical IOI 21 Independent Variable: Critical IOI

64 Rationale Rationale In a rhythmic context, attending is paced in time. Pitch judgments will be best for temporally expected tones: When critical IOI = context IOI.

65 Similar Predictions Aquadratic expectancy A quadratic expectancy profile Attentional Pulse

66 P<.005

67 Related Experiments Found a: flattened expectancy profile with irregular context timing

68 Recent findings Suggest the importance of higher-order time structure … In focal attending over time levels

69 An Accent Rhythm An Accent Rhythm 21 Accent Timing Accented Tones Higher time level : 4 IOIs Lower time level: IOI

70 Accent Rhythm is Varied Accent Rhythm is Varied 21 Regular Irregular Accented Tones

71 Standard pitch Comparison pitch Same Pitch Judgment Task Critical IOI is Varied 21

72 Two Oscillators? Remember those two circadian oscillators (Morning, Evening)? Well, these are different! Shorter periods Sensitive to sound

73 Two Different Oscillators Two oscillators : Different periods period ~ IOI. 1. Lower-order oscillator, period ~ IOI. period ~ 4 IOIs 2. Accent oscillator, period ~ 4 IOIs only If the Accent oscillator dominates, then only the Regular Rhythm will give an Expectancy Profile. Which dominates?

74 PC P<.001

75 Two Oscillators In listening to sound patterns Differential responses to sounds: Accented versus Unaccented Flexible attending: Dominance of one oscillator in focal attending.

76 Final Comments The dynamics of sound patterns together the dynamics of attending may explain real time tracking of events. Assumptions about rhythm and attentional synchrony can enhance theoretical power. Entrainment to dynamic contexts may play a role in judgments of both time and pitch.

77 Thank you ! Thank you !

78

79

80 Two routes to Synchrony 1. Anticipatory attending: Attending heightens prior to an element. Oscillator period is important. 2. Reactive attending: If anticipation fails, If anticipation fails, Attending shifts after an element occurs. Oscillator phase is important. Oscillator phase is important.


Download ppt "By Mari Riess Jones The Dynamics of Attending and Scene Analysis."

Similar presentations


Ads by Google