Gupta, Rishabh
(2016).
Physiology-based Quality-of-Experience Assessment for Next Generation Multimedia Technologies.
Thèse.
Québec, Université du Québec, Institut national de la recherche scientifique, Doctorat en télécommunications, 176 p.
Résumé
As new multimedia technologies emerge, telecommunication service providers have to provide
superior user experience in order to remain competitive. To this end, quality-of-experience (QoE)
perception modelling and measurement has become a key priority. QoE models rely on three influence
factors: technological, contextual and human. Existing solutions have typically relied on the
former two and human influence factors (HIFs) have been mostly neglected due to difficulty in measuring
them. In this thesis, we show that measuring HIFs is important for QoE measurement and
propose the use of hybrid brain-computer interfaces (hBCIs) for objective measurement of perceived
QoE for multimedia technologies, such as affective music videos and text-to-speech systems.
For the development of hBCIs, we explore the use of two neuroimaging techniques, namely electroencephalography
(EEG) and functional near-infrared spectroscopy (fNIRS), to better understand
neuronal and cerebral haemodynamic changes resultant from multimedia signals of varying quality.
Neural correlates of several QoE dimensions were derived and validated on the publicly available
DEAP and PhySyQX databases. In general, the parameters derived from EEG and fNIRS indicated
correlation between neural activation, in various cortical regions, and signal quality. These individual
features derived from EEG and fNIRS were then used to develop classifiers to establish their
usability as QoE monitoring modalities. The parameters derived from EEG and fNIRS showed to
accurately classify different user states and subjective QoE dimensions. Interestingly, features derived
from heart rate, extracted from fNIRS signals, also showed to encode information regarding
HIFs. Next, fusion of EEG, fNIRS, and fNIRS-derived heart rate parameters showed to accurately
represent several QoE dimensions, including those related to listener affective states.
Finally, the subjectively-derived HIFs were incorporated into the QoE model, leading to gains
of up to 26.3% relative to utilizing only technological factors. When utilizing HIFs derived from
individual modalities, on the other hand, gains of up to 14.5%, 10.6% and 4% were observed for EEG,
fNIRS and heart rate, respectively. The hybrid model based on features from all three physiological
modalities resulted in gains of up to 18.4%. These findings show the importance of using BCIs
and hBCIs in QoE measurement and also highlight that further improvement may be warranted
once improved HIFs correlates are found from EEGs and/or other neurophysiological modalities.
It is hoped that these findings will help researchers build better instrumental QoE models that
incorporate technological, contextual, and human influence factors.
Actions (Identification requise)
 |
Modifier la notice |