HUMAN COMPUTER HUMAN COMPUTER INTERACTIONSUBJECT CODE DCM 214SUBJECT CODE : DCM 214

Preppared by : NU

RRA

INI M

OH

D

Chapter 2HUMAN CAPABILITIES :

1

GH

AN

I

INPUT OUTPUT SYSTEMS

KEY POINT

Human have processing constraintsp gMotor limitations, e.g. Fitts’ law for pointingVisual range for motion, shape, colour, detail and

Prepare

their consequences for design decisionsVisual attention modelsAlt ti h l

ed by : NU

RAIAlternative sensory channels IN

I MO

HD

GHH

AN

I

2

HUMAN CONSTRAINTS

What do we know about human capabilities that could orshould constrain interface design?

Limits on perceptual capability – e.g. contrast, resolution

Preparep p p y g ,Limits on motor capability – e.g. reach, speed, precisionLimits on attention capacityLimits on memory

ed by : NU

RAILimits on memory

Rates of learning and forgettingCauses of errorM t l d l & bi

INI M

OH

D G

H

Mental models & biasesIndividual differences (the average size fits few people)Variable state (e.g. stress, fatigue)

HA

NI

Special needs & age … 3

HUMAN CONSTRAINTS

Model Human Processor(MHP)

*One way to subdivide

Prepareythe main constraints*Perceptual, Motor andCognitive sub-systems

ed by : NU

RAICognitive sub systems

characterised by:– Storage capacity U– Decay time D

INI M

OH

D G

HDecay time D– Processor cycle time T*We will focus today onthe perceptual and

HA

NI

the perceptual andmotor processes

4

MOTOR CONSTRAINTS

E l Fi ’ l (1954)

Prepare

Example: Fitts’ law (1954)

ed by : NU

RAIIN

I MO

HD

GHH

AN

I

5

MOTOR CONSTRAINTS

Example: Fitts’ law (1954)p ( )

Prepare

Justification?

ed by : NU

RAIJustification?

#By “analogy” to Shannon informationcapacity = bandwidthxlog2((signal+noise)/noise)

INI M

OH

D G

H

#If move fraction 1-r to target each timestep, then reach target when rnD = W/2; so n is proportional to log22D/W

HA

NI

proportional to log22D/W# Empirically find good fit with log2(D/W + 0.5)

6

MOTOR CONSTRAINTS

Example: Fitts’ law (1954)p ( )

Prepareed by : NU

RAI

Application?*Time will increase with distance – can we

keep everything close?

INI M

OH

D G

Hkeep everything close?*Time will decrease with width – can we

make width infinite?

HA

NI

7

PERCEPTION

What can we see?

Prepareed by : NU

RAIIN

I MO

HD

GHH

AN

I

8

PERCEPTION

Some consequences of what we can see:q#Motion – will be visible (and distracting)

anywhere in visual field

Prepare

# Colour – main advantage is “pop-out”:But many disadvantages:

# Sh i t t i t t iti SO

ed by : NU

RAI# Shape important in text recognition: SO

ALL CAPS BAD# Limits on resolution – recommend

INI M

OH

D G

H

minimum font size; ideally individual can adjust# High resolution only in tiny area of fixation

HA

NI

9

EYE TRACKING

Fixation pattern is a good indicator of attentionp g

#Where do people look, how often, for how

Prepare

long, in what order?#Recent technology is making this a

standard tool for HCI

ed by : NU

RAIstandard tool for HCI

#Also used as input device.

INI M

OH

D G

HHA

NI

10

PERCEPTION

Prepare

Importance of eye movements

Must shift the tiny high resolution area around ed by : N

UR

AI

resolution area aroundConstantly

Movements called saccades INI M

OH

D G

H

occur > 2 per second all day Long

How does visual system decidewhere to move next? H

AN

I

where to move next?

Models of attentione.g. Itti et.al. 1998

11

ATTENTIONSimple statistical model of saliency Rosenholtz et al (2005)

*Provides definition of ‘clutter’: size oflocal covariance ellipsoid

Prepare

* To measure:* Compute local feature covariance atmultiple scales

ed by : NU

RAI

* Take maximum across scales* Average for different features* Pool over space* P d d l ti ith h

INI M

OH

D G

H

* Produces good correlation with humanestimates of clutter* Can also use to determine whatfeature added where would best draw

HA

NI

feature added where would best drawattention

12

ATTENTION#So what went wrong here?

k fi d l i f S#Task: find current population of U.S.

Prepareed by : NU

RAIIN

I MO

HD

GH

86% of users failed…

HA

NI

http://www.useit.com/alertbox/fancy-formatting.html 13

PERCEPTUAL CONSTRAINTS

Bottom up visual processing sets some constraints on ti l l t b t t l id t d optimal layouts, but must also consider top down

issues:

#C lt l d l d f t f ili it

Prepare#Cultural and learned factors – familiarity#Underlying domain knowledge of user# Need to reflect logical structure, e.g., placement and

grouping

ed by : NU

RAIgrouping

according to function, sequence, frequency of use# Dependence on task to be carried out, e.g. getting an

overview

INI M

OH

D G

Hoverviewvs. seeking specific information# Note that layout and visualisation are already widely

explored fields with conclusions that carry over to

HA

NI

explored fields, with conclusions that carry over to HCI 14

ALTERNATIVE SENSORY CHANNELS

Different sensors provide parallel channel capacitySound:

#Not so easy to localise but can detect from any direction

Preparey y# Grabs attention – warning mechanisms# Good signal of causal relation – use as confirmatory

feedback

ed by : NU

RAI

# Monitoring state, ‘background information’# Disk, printer noise etc.# Example of user improvisation in use of ‘data’

INI M

OH

D G

H# Example of user improvisation in use of data# Interface sound design is typically arbitrary and synthetic

To ch and haptics:

HA

NI

Touch and haptics:# Exploit our natural ability to ‘handle’ objects

15

THANK YOUTHANK YOUSEE YOU NEXT CLASSSEE YOU NEXT CLASS

Prepare

AND

ed by : NU

RAI

DON’T FORGET TO

INI M

OH

D G

HDON T FORGET TO FINISH YOUR

HA

NI

HOMEWORK 16