Psychophysics ECE 51100
Popular in Course
Popular in Electrical Engineering & Computer Science
This 19 page Class Notes was uploaded by Cassidy Casper on Saturday September 19, 2015. The Class Notes belongs to ECE 51100 at Purdue University taught by Staff in Fall. Since its upload, it has received 36 views. For similar materials see /class/207908/ece-51100-purdue-university in Electrical Engineering & Computer Science at Purdue University.
Reviews for Psychophysics
Report this Material
What is Karma?
Karma is the currency of StudySoup.
Date Created: 09/19/15
Dimensionality Hong Z Tan amp Zygmunt Pizlo How to Achieve High IT I IT for unidimensional stimuli is limited I ITmultiD is not limited by 72 I In general try OLots of dimensions 9A few values 2 to 3 per dimension 0 Examples Speech perception Face recognition Hong Z Tan amp Zygmunt Pizlo How do you define dimensionality I From literature never explicitly de ned I Read between lines number of independently manipulated physical variables I But physical and perceptual dimensionality may not be the same Hong Z Tan amp Zygmunt Pizlo Dimensionality a Visual Example I Orientation of lines 1D or 2D I IT for direction or angle of inclination is 33 bits for a 5sec exposure time ref p 86 Miller s 72 paper I This is clearly at the high end of 72 23398 Hong Z Tan amp Zygmunt Pizlo Dimensionality an Auditory Example Lateralization A L V r Rough Two Clicks L 0 1 10 V1 msec Interaural Time Delay Hong Z Tan amp Zygmunt Pizlo Dimensionality a Haptic Example Motion Flutter Vibratian O O A A C O l l l l I I I lJ l I I l I I LlJ I l 2 4 10 30 150 300 Frequency Hz Hong Z Tan amp Zygm um Pizllo IT and Channel Capacity For Different Sensory Modalities I AL and DL are in modalityspeci c physical units I IT and channel capacity are in bits We can compare apples with oranges Hong Z Tan amp Zygmunt Pizlo Cumulative d39 and its Relationship to IT Hong Z Tan amp Zygmunt Pizlo Overview I Based on two papers Intensity perception I amp II by Durlach amp Braida Journal of the Acoustical Society of America 1969 amp 1972 I Goal Towards a theory for interpreting the relation between our ability to discriminate between two intensities that differ only by a small intensity increment and our inability to identify an intensity from among a large set of intensities that differ by very large increments the 7i2 phenomenon Hong Z Tan amp Zygmunt Pizlo The Formal Theory I Decision model I Internalnoise model 6 Quanti es M and ain the decision model in terms of sensory and memory noise 6 Sensory noise is mainly due to the subject s inability to maintain the image or the trace of the sensation precisely 0 Memory noise is due to the subject s inability to remember the general context of sounds in the experiment and the inability to determine or represent the relation of the sensation to this context precisely Hong Z Tan amp Zygmunt Pizlo Decision Model Revisited d39szM1 a Assume that Weber s Law holds then M Kloin i1 2 Assume that the variance 0392 is the sum of sensory noise 32 independent of D and memory noise GZR2 Where Gconstant RlogImaXImin ie 02 32 Gsz It follows that an K10g12I1 101211 M2 62122 M102 GKYRZ Hong Z Tan amp Zygmunt Pizlo Resolution in OneInterval Paradigms 10g1211 I d mode2122 12 Can be extended to a Wide variety of oneinterval experiments including the discrimination and absolute identi cation paradigms to measure the sensitivity index between any pair of stimuli Il and d3911 j Z 101139 1 3102 61021 Hong Z Tan amp Zygmunt Pizlo d in Al Experiments loZgI1 loigl2 10g1k X I Sensitivity index d39 is additive d39I311 d391211 d391312 ie 10g1311 10g1211 10g13 12 lt Ilt2ltGKgt2R2 lt Ilt2ltGIlt2R2 mama10W 13 Hong Z Tan amp Zygmunt Pizlo Cumulative d Links ll2AFC and AI Experiments I Cumulative d or total sensitivity can be expressed as 10g1k11 A39 Z d391max mi do 2 my GK2R2 or equivalently A d1max1min R W GK2R2 I Cumulative d is a function of R only I When R is large A39zKG ie constant 14 Hong Z Tan amp Zygmunt Pizlo Cumulative d39 vs R SF quotlt1 392 2 l0 0 a 1 2 j 01 5 E h o l i I1 I Jl l J 0 I 2 3 4 5 6 7 INTENSITY RMGE R I Circles experimental data I Curve derived with KG137 and KB81 Hong Z Tan amp Zygmunt Pizlo Predicting IT from A I Assumptions 9 The means of the density functions on the decision aXis are equally spaced 9 The number of responses equals the number of stimuli O The response criteria are placed midway between adjacent means I The stimulusresponse confusion matrix can be predicted Hong Z Tan amp Zygmunt Pizlo IT vs Cumulative d LO 939 a mum mmnmnnu ms 5 o r 7 4 0 tom stlsmvnrv a39 I Crosses one subject AI experiments with N10 I Curve theoretical prediction Hong Z Tan amp Zygmunt Pizlo TOTAL ssusmvm A39 A New Interpretation of 7i2 9 O 1 o MUTUAL NORMAN 0 III39I SI O l L I l 4 l J 0 I 2 3 4 5 5 INTENSITY RANGE R 14Lnll YOTIL SEISI I WITI39 A39 I Maximum A E 12 15 estimated from experimental data I Therefore IT for intensity is limited Hong Z Tan amp Zygmunt Pizlo Further Readings I N I Durlach and L D Braida Intensity perception 1 Preliminary theory of intensity resolution The Journal of the Acoustical Society of America vol 46 pp 372 383 1969 I L D Braida and N I Durlach Intensity perception 11 Resolution in oneinterval paradigms The Journal of the Acoustical Society of America vol 51 pp 483 502 1972 19 Hong Z Tan amp Zygmunt Pizlo