Samenvatting
Summary 0HM140 Advanced Perception
Summary on all lectures and background literature for the HTI-master course 0HM140 Advanced Perception at the TU/e.
[Meer zien]Voorbeeld 4 van de 54 pagina's
Voorbeeld 4 van de 54 pagina's
In winkelwagengesponsord bericht van onze partner
3 beoordelingen
Door: timodamoiseaux • 2 jaar geleden
Niet alle stof wordt besproken, terwijl het totaal niet beknopt is
Door: fleurdekorver • 2 jaar geleden
Heel veel onderwerpen worden niet besproken. Plus de onderwerpen die worden bessproken is niet beknopt genoeg
Door: darreman • 3 jaar geleden
Voorbeeld van de inhoud
0HM140 Advanced Perception University of Technology EindhovenLecture 1: Introduction to Psychophysics
Perception is the organization, identification, and interpretation of sensory information in order
to represent and understand the presented information, or the environment. All perception
involves signals that go through the nervous system, which in turn result from physical or
chemical stimulation of the sensory system. For example, vision involves light striking the retina
of the eye, smell is mediated by odor molecules, and hearing involves pressure waves. By
understanding how stimulus properties influence our view on the world, we are able to create
(via multimedia but also virtual reality) new environments and improve existing environments.
Perception is not always accurate and veridical. It is important to understand these
misjudgments and what can be done to avoid them. Perceptual systems have limited
capabilities in detection a stimulus. This defines the accuracy of interfaces (e.g. the resolution
of displays). Capabilities of artificial intelligence systems are nowhere near human perception
(e.g. speech, face, object recognition). We can learn from human perception to improve their
performance. There are several ways through which we can study the senses:
Behavioral studies
o Introspection (no unconscious processes): the examination of one’s own
conscious thoughts and feelings and relies solely on own observations.
o Phenomenology (inherently subjective; qualitative): the study of “phenomena”
as in appearances of things, or things as they appear in our experience, or the
ways we experience things, thus the meanings things have in our experience.
o Psychophysics (purely functional): field in psychology that particularly concerns
the way we perceive the world which involves quantitative measurements of
psychological phenomena. Done in several steps: present stimuli, measure
response, infer internal and psychological quantities. As such, the
psychophysical linking hypothesis proposes a specific causal link between neural
activity in the brain and perceptual experience. Psychophysics quantitatively
investigates the relationship between physical stimuli and the sensations and
perceptions they produce. Psychophysics has been described as "the scientific
study of the relation between stimulus and sensation" or, more completely, as
"the analysis of perceptual processes by studying the effect on a subject's
experience or behaviour of systematically varying the properties of a stimulus
along one or more physical dimensions". In general, psychophysics is
quantitative (objective scale of measurement), does not suffer from subjectivity
or introspection, can be used to study “pure” mental phenomena (e.g.
attention) and valid inter-subject and inter-method comparisons, can identify
(subconscious) response bias and takes into account internal noise.
Neurophysiological studies
o Lesion studies: removal of tissue and observe change in alternate behavior.
Page | 1
,0HM140 Advanced Perception University of Technology Eindhoven
o Single cell recordings: insert electrode in the brain and measure changes in
electrical currents after presentation of a stimulus. Though it’s crucial for
understanding individual neurons, it’s very hard to generalize huge populations.
o Histology: study of the microscopic anatomy of cells and tissues of organisms.
o Brain imaging: structural imaging is based on difference in physical properties of
tissues (e.g. computerized tomography CT which relies on tissue density and
magnetic resonance imagine MRI which relies on absorption of X-rays).
Functional imaging is based on the assumption that neural activity produces
local physiological changes in the region of the brain (e.g. Positron Emission
Tomography scans measure blood flow in a region and functional MRI measures
blood oxygen levels in BOLD responses). Normally, structural imaging has a bad
temporal resolution while functional imaging has a bad spatial resolution.
o Transcranial Magnetic Stimulation: temporarily disrupt cognitive function using
direct electrical stimulation to create a virtual lesion.
Theoretical/modelling
o Computational perception models: the study of the computational basis of
learning and inference by mathematical modeling, computer simulation, and
behavioral experiments. In psychology, it is an approach which develops
computational models based on experimental results. It seeks to understand the
basis behind the human method of processing of information.
o Computational neuroscience: the field of study in which mathematical tools and
theories are used to investigate brain function. It can also incorporate diverse
approaches from electrical engineering, computer science and physics in order
to understand how the nervous system processes information.
There are several problems that may occur when studying the senses. First of
all, experiences are personal and sensations differ from person to person. Also,
sensory processes do not always result in veridical representations of the
physical world (Muller-Lyer illusion), sometimes due to the fact that experiences are personal
and perceptions differ from person to person. Mapping is usually not 1 to 1 because of the fact
that the same stimulus can give rise to different responses while different stimuli can still result
in the same sensation. Also, responses are noisy since people give different responses to the
same stimulus over time (e.g. because of fatigue, strategy, distraction). The conclusion is that
responses are probabilistic and all of this makes it difficult to, with help of a stimulus variable
(e.g. RGB color) and a dependent variable (e.g. matched color by participant), get information
of internal variables (perceived color). There various psychophysical methods that try to
measure the observer’s threshold while still considering these bias problems:
Let people rate magnitude of a given stimulus: (e.g. distance between objects or
loudness of a sound). No reference stimulus is given and the participant just rates one
stimulus. This is subjective and not very accurate because participants do not agree on
Page | 2
,0HM140 Advanced Perception University of Technology Eindhoven
units and are bad at making absolute judgments. When participants are asked to rate
magnitude on a given scale (e.g. 1-10), there is a reference scale though for some stimuli
this is not always possible and changing the scale changes the results.
Doing matching experiments: a presented stimulus needs to be rated on its magnitude
on a given scale. Even though this is better than not providing a reference at all, a good
reference scale is not always possible and it only involves relative judgments. An
example of a matching experiment is the method of adjustment, explained later.
Measure psychometric function: for example, the forced-choice task which is a
psychophysical procedure in which each representation contains two intervals, only one
of which contains the stimulus (randomly selected). The subject must select the correct
interval. There are only 2 choices (e.g. ‘was there a sound on the left or right?’) Another
example is the yes/no task in which only 50% of presentations contain a stimulus and
the subject must respond ´yes’ or ‘no´ after each. Both these tasks reduce the possibility
of bias because when the stimulus is not detectable, the subject is forced to guess as to
which event contained the stimulus.
Psychometric functions are generally very accurate, only functional, measure mental states,
suitable for measuring detection thresholds (the smallest magnitude that can be perceived) and
discrimination thresholds (the smallest difference that can be perceived) though they require
many measurements. We define the point of subjective equality as the stimulus intensity for
which the chance of detection is 50% (discrimination threshold). There are a few commonly
used methods:
Methods of adjustment: subjects adjust stimulus level until they can just
detect/discriminate it. This type of method directly measures the threshold and is a
quick way of determining correct stimulus ranges though subjects can be biased (e.g.
reluctant to say ‘no’). Therefore, there is relatively much internal noise.
Methods of limits: stimulus level is either
incremented until the signal is detected/
discriminated or decreased until the signal is
no longer detected / discriminated (absolute
detection threshold). The output of such a
method is thus two estimates and prevents subjective bias (take the average threshold).
Furthermore, we distinguish between errors of habituation (stimulus is likely to be the
same) and errors of anticipation (stimulus is likely to be different). To get rid of
habituations and anticipation, one should use the method of constant stimuli.
Page | 3
, 0HM140 Advanced Perception University of Technology Eindhoven
Methods of constant stimuli: measures the probability of
detection for several predefined stimulus levels which are
constant. This method gives information about both
sensitivity and threshold though in general it is more
accurate that the other methods. In this case, the
sensitivity is the inverse of steepness (e.g. steeper for light
than for sound). Basically, the sensitivity is the inverse of
the threshold. Pretests are necessary for choice of range and
many measurements are needed. So after measuring the
probability of detection for a number of stimuli, you fit a
psychometric function to it. The method of constant stimuli
thus gives an estimation of both slope and threshold (50%)
and is quite accurate. Pretests are necessary for choice of
range and it requires many trails (> 10 per stimulus level). Participants can have a
different criterion (bias) for their threshold (reluctant or liberal).
Adaptive methods: to speed things up. Adaptive
methods can be optimized for estimating the
threshold only, or threshold and slope. Adaptive
methods are classified into staircase procedures and
Bayesian or maximum-likelihood methods. Staircase
methods rely on the previous response only and are
easier to implement. Stimulus intensity depends on
the response to a previous stimulu. Examples: the
stimulus is reduced after two “yes” responses by one step size; the stimulus is increase
after one “no” responses by one step size; the threshold is the mean of the reversal
points.
However, problems with a yes/no task is that as an
experimenter, you have no control about response
bias (how likely is the participant to say yes or no), the
attention of the participant, memory of the
instruction and how serious the participant
participates. Thus it is important to separate
sensitivity and response bias. The signal detection
theory acknowledges the importance of bias effects by assuming that stimulus detection is a
two-stage process. The first stage is a purely sensory process in which a specific stimulus level
produces an internal sensory response that depends on the intensity of the stimulus and the
sensitivity of the sensory system. The second stage is a decision process in which the sensory
response magnitude is compared to an internally set criterion. If the response magnitude
exceeds this criterion, the decision process decides that a stimulus was present. There are a
number of possible responses: correct detections, missed detections, correct rejections and
Page | 4
Voordelen van het kopen van samenvattingen bij Stuvia op een rij:
Verzekerd van kwaliteit door reviews
Stuvia-klanten hebben meer dan 700.000 samenvattingen beoordeeld. Zo weet je zeker dat je de beste documenten koopt!
Snel en makkelijk kopen
Je betaalt supersnel en eenmalig met iDeal, creditcard of Stuvia-tegoed voor de samenvatting. Zonder lidmaatschap.
Focus op de essentie
Samenvattingen worden geschreven voor en door anderen. Daarom zijn de samenvattingen altijd betrouwbaar en actueel. Zo kom je snel tot de kern!
Veelgestelde vragen
Wat krijg ik als ik dit document koop?
Je krijgt een PDF, die direct beschikbaar is na je aankoop. Het gekochte document is altijd, overal en oneindig toegankelijk via je profiel.
Tevredenheidsgarantie: hoe werkt dat?
Onze tevredenheidsgarantie zorgt ervoor dat je altijd een studiedocument vindt dat goed bij je past. Je vult een formulier in en onze klantenservice regelt de rest.
Van wie koop ik deze samenvatting?
Stuvia is een marktplaats, je koop dit document dus niet van ons, maar van verkoper brampeters. Stuvia faciliteert de betaling aan de verkoper.
Zit ik meteen vast aan een abonnement?
Nee, je koopt alleen deze samenvatting voor €6,49. Je zit daarna nergens aan vast.
Is Stuvia te vertrouwen?
4,6 sterren op Google & Trustpilot (+1000 reviews)
Afgelopen 30 dagen zijn er 62491 samenvattingen verkocht
Opgericht in 2010, al 14 jaar dé plek om samenvattingen te kopen