EE Audio Signals and Systems

Slides:



Advertisements
Similar presentations
Adaptive Methods Research Methods Fall 2010 Tamás Bőhm.
Advertisements

Signal Detection Theory. The classical psychophysicists believed in fixed thresholds Ideally, one would obtain a step-like change from no detection to.
THE DISTRIBUTION OF SAMPLE MEANS How samples can tell us about populations.
M. Zareinejad.  Methodology for investigating relationships between sensations in the psychological domain and stimuli in the physical domain  Central.
Chapter 10: Estimating with Confidence
PHYSICS OF SOUND PHYSICS OF SOUND HEARING CONSERVATION PROGRAM 1 28 Jan 2013.
Thresholds, Weber’s law, Fechner’s three methods Research Methods Fall 2010 Tamás Bőhm.
PSYCHOPHYSICS What is Psychophysics? Classical Psychophysics Thresholds Signal Detection Theory Psychophysical Laws.
Rob van der Willigen designed by Stephanie Thái.
9-1 Hypothesis Testing Statistical Hypotheses Statistical hypothesis testing and confidence interval estimation of parameters are the fundamental.
8-1 Quality Improvement and Statistics Definitions of Quality Quality means fitness for use - quality of design - quality of conformance Quality is.
Rob van der Willigen designed by Stephanie Thái.
Today Concepts underlying inferential statistics
Chapter 10: Estimating with Confidence
EE Audio Signals and Systems Psychoacoustics (Masking) Kevin D. Donohue Electrical and Computer Engineering University of Kentucky.
Chapter 9 Title and Outline 1 9 Tests of Hypotheses for a Single Sample 9-1 Hypothesis Testing Statistical Hypotheses Tests of Statistical.
EE513 Audio Signals and Systems Statistical Pattern Classification Kevin D. Donohue Electrical and Computer Engineering University of Kentucky.
EE513 Audio Signals and Systems Digital Signal Processing (Systems) Kevin D. Donohue Electrical and Computer Engineering University of Kentucky.
Applied Psychoacoustics Lecture 4: Loudness Jonas Braasch.
Learning Objectives Copyright © 2004 John Wiley & Sons, Inc. Sample Size Determination CHAPTER Eleven.
+ The Practice of Statistics, 4 th edition – For AP* STARNES, YATES, MOORE Chapter 8: Estimating with Confidence Section 8.1 Confidence Intervals: The.
9-1 Hypothesis Testing Statistical Hypotheses Definition Statistical hypothesis testing and confidence interval estimation of parameters are.
Physics 114: Exam 2 Review Lectures 11-16
COSC 1P02 Introduction to Computer Science 4.1 Cosc 1P02 Week 4 Lecture slides “Programs are meant to be read by humans and only incidentally for computers.
Chapter 7: Loudness and Pitch. Loudness (1) Auditory Sensitivity: Minimum audible pressure (MAP) and Minimum audible field (MAF) Equal loudness contours.
L Berkley Davis Copyright 2009 MER301: Engineering Reliability Lecture 9 1 MER301:Engineering Reliability LECTURE 9: Chapter 4: Decision Making for a Single.
Fundamentals of Sensation and Perception EXPLORING PERCEPTION BY STUDYING BEHAVIOUR ERIK CHEVRIER SEPTEMBER 16 TH, 2015.
EE Audio Signals and Systems Linear Prediction Kevin D. Donohue Electrical and Computer Engineering University of Kentucky.
Psychophysics and Psychoacoustics
Spatial Smoothing and Multiple Comparisons Correction for Dummies Alexa Morcom, Matthew Brett Acknowledgements.
Response Processes Psych DeShon. Response Elicitation Completion Completion Requires production Requires production Allows for creative responses.
Absolute Threshold PSY 3215 – Perception – Appalachian State University.
Outline of Lecture I.Intro to Signal Detection Theory (words) II.Intro to Signal Detection Theory (pictures) III.Applications of Signal Detection Theory.
Psy Psychology of Hearing Psychophysics and Detection Theory Neal Viemeister
Signal detection Psychophysics.
+ The Practice of Statistics, 4 th edition – For AP* STARNES, YATES, MOORE Chapter 8: Estimating with Confidence Section 8.1 Confidence Intervals: The.
Chapter 8: Estimating with Confidence
Chapter 8: Estimating with Confidence
Physics 114: Exam 2 Review Weeks 7-9
Methods in Brain Research: psychophysics
What type of wave is this?
Sensation: Psychophysics
"Digital Media Primer" Yue-Ling Wong, Copyright (c)2013 by Pearson Education, Inc. All rights reserved.
EE Audio Signals and Systems
Signal Detection Theory
EE Audio Signals and Systems
Classical psychophysics
9 Tests of Hypotheses for a Single Sample CHAPTER OUTLINE
PSY Perception – Appalachian State University
Signal Detection Theory
EE513 Audio Signals and Systems
Chapter 8: Estimating with Confidence
Chapter 8: Estimating with Confidence
Chapter 8: Estimating with Confidence
Chapter 8: Estimating with Confidence
Signal detection theory
Chapter 8: Estimating with Confidence
Chapter 6 (B): Thresholds and Sensory Adaptation
ECEN 615 Methods of Electric Power Systems Analysis
Chapter 8: Estimating with Confidence
Chapter 8: Estimating with Confidence
PSY 250 Hunter College Spring 2018
Chapter 8: Estimating with Confidence
Chapter 8: Estimating with Confidence
Chapter 8: Estimating with Confidence
Chapter 8: Estimating with Confidence
Chapter 8: Estimating with Confidence
HKN ECE 313 Exam 2 Review Session
Principles of perception
Volume 23, Issue 11, Pages (June 2013)
Presentation transcript:

EE599-020 Audio Signals and Systems Psychoacoustics Kevin D. Donohue Electrical and Computer Engineering University of Kentucky

Psycho Psychoacoustics – study of the sensations produced by sounds. Psychophysics – study of relationship between magnitude of stimulus and magnitude of sensory response or sensation.

Psychometric Function A particular human response to a stimulus can be considered a signal. A psychometric function maps a stimulus measure into a probability that a particular human response will occur.

Psychometric Experiment Response: Human hears a change in pitch Stimulus: A change in the tonal frequency with measure being the frequency difference between the tone in Hertz Design an experiment to estimate the psychometric function indicating the smallest tonal interval from 1000 Hz that can be detected by a human. Run experiment using a yes/no test and plot the estimated psychometric function.

Response Model of Yes-No Test Assume human has an internal axis r on which the stimulus is mapped to a response (recall definition of random variable!). Assume a stimulus will map to this internal axis according to a Gaussian distribution with mean (  ) proportional to the stimulus measure (strength) and standard deviation (  ) representing intra- and inter-subject variability of the responses.

Internal Decision Criterion Consider the types of errors for user response and derive expressions for their the probabilities.

Compute Decision Criterion The response for a given stimulus is modeled by a Gaussian distribution with mean 6 and standard deviation 1, find the decision criterion threshold where the human would have the correct response 92% of the time. If “no stimulus” is modeled by the same distribution with a mean of zero, what is the false alarm (false positive) rate? Hint: Use the inverse CDF, also use Z-score transformation and standard error function (inverse).

Limitations with Yes/No Tests The Yes/No test involves the internal criteria that subjects use to decide whether the stimulus is present. For subjects with similar low level auditory functions this criteria can vary depending on how liberal (aggressive) or conservative the subject is in making positive decisions. It can even vary in the same subject depending on mood or concentration level. A 2 alternative forced choice test is typically used reduce the effects of the various internal criteria and emphasize the auditory systems detection capability. Describe the difference in false positive errors between aggressive and conservative subjects in a Yes/No test?

2AFC Test The 2 alternative forced choice (2AFC) tests present a subject with 2 intervals. One contains the stimulus and the other does not. The subject must decide (guess in cases where they don’t know) which interval contains the stimulus. The resulting psychometric function plots strength of stimulus measure vs. probability of choosing the correct interval.

2AFC Test What is the probability of a correct response when the subject is guessing in a 2AFC?

2 AFC Trials - Staircase Method For a typical psychoacoustic the test stimuli are presented in sequence with increasing or decreasing values depending on the accuracy of subject response. A standard staircase method uses equal increments and decrements in the stimulus measure with turn around rules designed to converge on a specific probability threshold. The convergence probability is where the probability of turning down is equal to the probability of turning up (.5).

2 AFC Trials – Staircase Rules The “1-up n-down” rule operates to increase the stimulus one unit after one incorrect decision and to decrease the stimulus one unit after n consecutive correct decisions. The convergence probability is determine from the place where the turn down probability (n consecutive correct decisions) is .5:

2 AFC Trials – Staircase Rules What is the convergence probability for a 1-up 2-down rule? What are all possible sequences that result in the staircase sequence turning up? What is the probability of it turning up? What are all possible sequences that result in the staircase sequence turning down? What is the probability of it turning down?

2 AFC Test for Loudness Modify the mfile for the yes/no pitch example to test subjects with a 2 AFC test to estimate the internal threshold on how much signal power (in dB) increase results in deciding when a test tone of 1000 Hz is louder than a reference tone of 1000 Hz at 75 dB. Use a 1-up 4-down rule. Note: Unless you have a SPL meter you cannot determined where 75 dB is. However this adjustment can be done with the speaker volume and should not affect the mfile. You will need to use the “sound” command rather than the “soundsc” command in Matlab so you can adjust the amplitude of the sine wave and have that directly impact the change in volume. Volume change should happen in equal increments of dB.

Homework Problems For the staircase rule 1-up and 3-down, indicate all possible decision sequences that would cause the staircase to go up and all possible sequences that result in the sequence going down (let I indicate incorrect decision and C indicate correct decision). Compute the convergence probability in this case verify that the probabilities of the up and down turnarounds are both .5.