Measuring motor actions and psychophysiology for task difficulty estimation in human-robot interaction

fencinghuddleΤεχνίτη Νοημοσύνη και Ρομποτική

14 Νοε 2013 (πριν από 3 χρόνια και 6 μήνες)

75 εμφανίσεις

Measuring motor actions and
psychophysiology for task difficulty
estimation in human-robot interaction
Domen Novak, Matjaž Mihelj, Jaka Ziherl,
Andrej Olenšek, Marko Munih
Human-robot interaction
Measuringhuman psychology
most robots equipped with sensitive force
and movement sensors
much more difficult to measure the user’s
subjective feelings
: stress, engagement…
possible approach:
combine motor actions and
psychophysiology
Psychophysiological measurements
measure how physiological
processes
such as heart rate and skin conductance
are affected by psychological
states
respirationheart rate
skin temperature skin conductance
Study goal
Our goal: combine measurements of
motor actions with psychophysiology
to
obtain a more accurate estimate of the
user’s preferences
The scenario in action
7 difficulty levels: ball becomes progressively smallerand faster
Measurement protocol
user performs task
for 2 minutes
forces, movements and psychophysiology
are measured
user asked if he/she would prefer easier
or hardertask
difficulty changed, task continues
(total 12 minutes)
Classification
Skin conductance analysis
Feature extraction
heart rate
mean HR
SDNN
RMSSD
pNN50
LF/HF ratio
HF power
LF power
skin
conductance
skin conductance
response frequency
mean skin conductance
response amplitude
final skin conductance level
skin
temperature
final temperature
respiration
mean respiratory rate
respiratory rate variability
task
performance
percentage of balls caught
percentage of balls in basket
difficulty level (1-7)
time since start
movement
sensors
mean absolute velocity
mean absolute acceleration
mean frequency
of position signal
mean frequency
of velocity signal
force sensors
mean absolute force
total work
mean frequency
of force signal
Motor actionsPsychophysiology
Linear discriminant analysis
a statistical method that finds an linear
boundary between two classes in a
multidimensional space
Adaptive discriminant analysis
major inter-individual differences,
especially in psychophysiology
requirement
: adaptationto the user
solution
: adaptive discriminant analysis,
where the discriminant function is
recursively updated online using Kalman
filtering
The classification process
Results
measure of success: percentage of times
the system can estimate the user’s
preferences
(prefer easier/ prefer harder)
participants:
24
healthysubjects
11
hemiparetic patients
leave-one-out crossvalidation
Results –healthy subjects
50
60
70
80
90
100
motor actionspsychophysiologyboth
Accuracy rate (%)
nonadaptive
adaptive
Results –patients
50
60
70
80
90
100
motor actionspsychophysiologyboth
Accuracy rate (%)
nonadaptive
adaptive
Conclusions
adaptive methods improve classification
of psychophysiological measurements
combination of motor actions and
psychophysiology offers small
improvement
over only motor actions
potentially more applicable to physically
undemanding
environments
Acknowledgements
Thank you for your attention
Research work was supported by
Research of the MIMICS project is funded through the
European Union under the 7th Framework programme
grant 215756
MOOG FCS kindly loaned one of two HapticMaster devices
for MIMICS