In computing and communications systems, quality is often difficult to define. Attempts to understand this concept date back to Aristotle, who included quality as one of his 10 categories of human apprehension. ISO standard 8402:1986 defines quality as “the totality of features and characteristics of a product or service that bears its ability to satisfy stated or implied needs,” which embraces objective as well as subjective parameters. In practice, however, quality could be compared to the elephant in the famous Indian parable about a group of blind men who each feels a different part of the animal and, thus, they disagree as to what it looks like....
3. Fernando Pereira: About Me …
• Associate Professor at University of Lisbon, Portugal
• Senior Researcher at Instituto de Telecomunicações, Lisbon, Portugal
• More than 250 publications in international journals and conferences
• One of the designers of the MPEG-4 and MPEG-7 standards
• ICIP, PCS, VCIP, WIAMIS, QoMEX General or Technical Program Chair
• Associate Editor of several journals
• Editor-in-Chief of the IEEE Journal of Selected Topics in Signal Processing (2013-2015)
• ISO/IEC Award for contributions to the MPEG-4 Visual Standard
• SPS Distinguished Lecturer
• IEEE Fellow in 2008 for “contributions to object-based digital video representation
technologies and standards”
• EURASIP Fellow in 2013 for “contributions to digital video representation technologies
and standards”
• IET Fellow in 2015
• IEEE SPS Board of Governors and EURASIP Board of Directors
• Several Excellence Teaching Awards
• JPEG (currently) and MPEG (past) Requirements Chair
8. Quality: Some Definitions from the Dictionary (1)
• Definition 1
– General : Measure of excellence or state of being free from
defects, deficiencies, and significant variations.
– ISO 8402-1986 standard defines quality as “the totality of
features and characteristics of a product or service that
bears its ability to satisfy stated or implied needs”.
• Definition 2
– Manufacturing : Strict and consistent adherence to
measurable and verifiable standards to achieve uniformity
of output that satisfies specific customer or user
requirements.
9. Quality: Some Definitions from the Dictionary (2)
• Definition 3
– Objective : Measurable and verifiable aspect of a thing or
phenomenon, expressed in numbers or quantities, such as
lightness or heaviness, thickness or thinness, softness or
hardness.
• Definition 4
– Subjective : Attribute, characteristic, or property of a thing
or phenomenon that can be observed and interpreted, and
may be approximated (quantified) but cannot be
measured, such as beauty, feel, flavor, taste.
13. What is QUALINET ?
• Group of institutions and companies interest in multimedia quality.
• Coordination of multidisciplinary QoE research in Europe and beyond.
• Strengthening dissemination efforts through already established, and new
initiatives, e.g. QoMEX, special events, books, journals.
• Strengthening interaction between academia and industry (industrial
forum, STSM, …).
• Strengthening educational efforts in QoE, e.g. summer schools, PhD
events, exchange of young researchers.
• Coordinated contribution to international standardization bodies, e.g.
ISO/IEC, ITU-T, VQEG, MPEG, JPEG.
In summary, group of researchers interested in (multimedia) QoE
issues, both theoretical and practical …
Open to new researchers … http://www.qualinet.eu/
19. QoS in Computer Networks and Communications
• Quality of Service (QoS)
– Resource reservation control mechanisms
– Ability to provide different priority to different
applications, users, or data flows
– Guarantee a certain level of performance (quality) to a
data flow
• (Service) Provider-centric concept
20. QoS Boundaries
• Scope: QoS typically focuses on telecommunications services.
• Focus: QoS deals with performance aspects of physical
systems.
• Methods: QoS has a very technology-oriented approach, and it
relies on analytic approaches and empirical or simulative
measurements.
from “Qualinet White Paper on Definitions of Quality of Experience”, March 2013
23. What is Mean Opinion Score (MOS)?
• Widely used in many fields:
– Politics/Elections
– Marketing/Advertisement
– Food industry
– Multimedia
– …
• The likely level of satisfaction of a specific service/product
dimension, e.g. visual quality, as appreciated by an averageuser
(from a provider point of view).
• Should be performed such that it generates reliable and
reproducible results
– Subjective evaluation methodology
– More complex and difficult that it a priori seems
– Much used for (and limited to) video and audio subjective qualities
25. FR, RR and NR Scenarios
• Full Referenceapproach:
• Reduced Reference approach:
• No-Reference approach:
Input/Reference
signal
Output/Processed
signal
Signal
processing
Input/Reference
signal
Output/Processed
signal
Signal
processing
FR METRIC
NR METRIC
Input/Reference
signal
Output/Processed
signal
Signal
processing
Features
extraction
RR METRIC
36. Multimedia Nowadays …
• Multimediais about sharing experiences
(real or imaginary) with others.
• In a way, it all started with story telling and
wall drawing around the fire in the caves
of early men.
• Modern multimedia systemsare evolved
versions of the good old story tellingand
wall drawing, which hopefullyoffer
increasingly richer experiences.
• The degree of richness of the experience
may be measured by Quality of Experience
(QoE).
38. What do People Talk about when they Talk about QoE ?
• “The degree of fulfillment of an intended experienceon a given
user.”
by Touradj Ebrahimi, 2001
• “perceived user experienceis psychological in nature and changes in
different environmental conditions and with different multimedia
devices.”
from QoMEX 2009 Call for Papers
• “The overall acceptability of an application or service, as perceived
subjectively by the end user.”
as defined by the ITU
The term ‘experience’ is appealing because it implicitly
promises individual engagement … Look good, sound good,
and feel good !
39. QUALINET QoE Definition
• Quality of Experience (QoE) is the degree of delight or
annoyance of the user (persona) of an application or service. It
results from the fulfillment of his or her expectations with
respect to the utility and/or enjoyment of the application or service
in the light of the user’s personality and current state (context).
• Experience: An experience is an individual’s stream of perception and
interpretation of one or multiple events.
• QoE feature: A perceivable, recognized and namable characteristic of the
individual’s experience of a service which contributes to its quality.
In the context of communication services, QoE can be influenced by factors
such as service, content, network, device, application, and context of use.
from “Qualinet White Paper on Definitions of Quality of Experience”, March 2013
40. Moving to Quality of Experience
• Quality of Service: Value of the average user’s service
richness estimated by a service/product/contentprovider
• Quality of Experience: Value (estimated or actually measured)
of a specific user’s experience richness
Quality of Experience is the dual (and extended) view of
Quality of Service !
QoS=provider-centric
QoE=user-centric
44. QoE Modeling
QoE modelingmay consider more or less
influence factors dependingon the
service/application, each with a different
weight on the overall assessment.
QoE is multi-dimensional, multi-modal and
multi-sensorial.
User centered influence factors are expected
to be dominating.
• System factors
– technical properties (as in
QoS)
• Human/User factors
– individual properties
– sensorial properties
– perceptual properties
– emotional properties
• Context factors
– environmental/physical
properties
– temporal properties
– service properties
– economic properties
– social properties
• Content factors
• …
50. QoS versus QoE
• Scope: QoS typically focuses on telecommunications services, whereas
QoE covers a much broader domain, which sometimes does not even
involve telecommunications.
• Focus: QoS deals with performance aspects of physical systems,
whereas QoE deals with the users' assessment of system
performance, as colored by context, culture, the users' expectations
with respect to the system or service and their fulfillment, socio-
economic issues, and psychological profiles, among other factors.
• Methods: QoS has a very technology-oriented approach, whereas QoE
requires a multi-disciplinary and multi-methodological approach for its
understanding.
• But it is also important to remember that QoE is, in a large part of
instances, highly dependent on QoS.
from “Qualinet White Paper on Definitions of Quality of Experience”, March 2013
56. QoE is Becoming Inevitable …
• Digital world has (re-)discovered the notion of quality
– Lower quality content is less and less tolerated by end-users in
some environments
– However, other environments seem to accept much lower
quality and still be successful
• Increasing interest in QoE
– Extending from device-centric and system-centric quality
optimization to end-to-end and especially user-centric
optimization
58. NOS UMA: an Example
• Ultra HD 4K
• Portability accross terminals, i.e. follows you
• Voice control (voice recognition ?), i.e. recognizes you
• User profiles within same family, i.e. individualizes you
• Recommendations based on user characterization, i.e.
targets you
• Complementary content for the favourite series, i.e. thinks
on you
• Time warping, i.e. helps you
• ...
A TV that
knows you !
59. QoE in Industry
• QoE is becoming mainstream.
• Many companies now speak about QoE.
• Personalization, interaction and recommendation capabilities
empower the user to create more individual experiences!
• However, QoE has a budget impact in terms of network and
system design, dimensioning, operation, maintenance, etc.
• But QoE is becoming more affordable in many application
domains …
• Embracing QoE principles may bring revenue, e.g. by
increasing viewing times and reducing churn.
60. Challenges Ahead
• Content-dependentqualityassessmentmethods and metrics
• Context-dependentqualityassessmentmethods and metrics
• Quality assessmentmethods and metrics beyond AV (haptics, smell,
…)
• Multi-modal quality assessmentmethods and metrics (AV, …)
• 3D quality assessmentmethods and metrics (3D sound, 3D video, …)
• New modalities content quality assessmentmethods and metrics
• Interaction qualitymetrics (closely related to usability)
• Presence/immersion qualitymetrics
• Role of emotions
• Virtual reality immersive experiences
• …
70. Take-Home Messages
• QoE is user-centric !
• QoE is individual, multidimensional and multisensorial.
• Services and systems are increasingly designed to
allow the users to maximize its QoE.
• Industry is increasingly embracing QoE principles
because they may bring revenue.
• QoE assessment is costly and risky but worth doing it.
• Int’l Conference on Quality of Multimedia Experience
(QoMEX): http://qomex.org/
72. Christian Timmerer: About Me …
• Associate Professor at Alpen-Adria-Universität Klagenfurt, Austria
(blog.timmerer.com, dash.itec.aau.at)
• Chief Innovation Officer | Head of Standardization and Research at
Bitmovin Inc., bitmovin.com
• Geschäftsführer Förderverein Technische Fakultät, ftf.or.at
• Lecturer | Carinthia University of Applied Sciences, www.fh-kaernten.at
• Research interest: immersive multimedia communication, streaming, adaptation, Quality
of Experience, and sensory experience
• More than 170 publications in international journals and conferences
• General chair: WIAMIS’08, QoMEX’13, QCMan’14, MMSys’16
• Associate editor/editorial board: IEEE Computer, IEEE Trans. on Multimedia, Signal
Processing: Image Communication, MTAP, IEEE Computing Now, ACM SIGMM Records,
ACM SIGMM OSSC
• Vice chair of IEEE ComSoC MMTC, WG leader in QUALINET
• Research projects: FP6-IST-DANAE (2004-2006), FP6-IST-ENTHRONE (2006- 2008), FP7-
ICT-P2P-Next (2008-2012), FP7-ICT-ALICANTE (2010-2013), FP7-ICT-SocialSensor (2010-
2014), COST-IC1003-Qualinet (2010-2014), FFG-AdvUHD-DASH (2014-2016), and FP7-ICT-
ICoSOLE (2013-2016)
• MPEG: MPEG-21, MPEG-M, MPEG-V, MPEG-DASH
• IEEE Senior member; ACM member
80. Scope of DASH: what is specified?
July 2016 ICME 2016 Tutorial, C. Timmerer 8
Media Presentation on
HTTP Server
DASH-enabled ClientMedia Presentation
Description
.
.
.
Segment
…
.
.
.Segment
…
.
.
.
Segment
…
.
.
.Segment
…
…
Segments located
by HTTP-URLs
DASH Control Engine
HTTP/1.1
HTTP
Client
MPD
Parser
Media
Engine
On-time HTTP
requests to
segments
81. Scope of DASH: what is specified?
July 2016 ICME 2016 Tutorial, C. Timmerer 9
Media Presentation on
HTTP Server
DASH-enabled ClientMedia Presentation
Description
.
.
.
Segment
…
.
.
.Segment
…
.
.
.
Segment
…
.
.
.Segment
…
…
Segments located
by HTTP-URLs
DASH Control Engine
HTTP/1.1
HTTP
Client
MPD
Parser
Media
Engine
On-time HTTP
requests to
segments
83. July 2016 ICME 2016 Tutorial, C. Timmerer 11
type=static typically,
for on demand content
Base URL of the
segments
Subtitles
Audio adaptation set
with different
representations (bw)
Video adaptation set
with different
representations (bw)
Different codecs
(profiles)
Segment URL constructed
with template and base
URL
85. Adaptive Streaming Content Workflow
July 2016 ICME 2016 Tutorial, C. Timmerer 13
Source Transcoding Encapsulation Encryption
Origin
Server
HelperDistribution
Client
Linear: Multicast
VoD: FTP, RTMP, HTTP, etc.
Unicast HTTP (PUSH), FTP, etc.
HTTP GET small objects
Single highest-bitrate
stream
Multiple streams at
target bitrates
Multiple streams at
target encapsulation formats
Large video/virtual
files and manifests
88. !
h t t p s : / / b i t m o v i n . c o m /
QoE Evaluation for DASH-based Services
• Test sequence
– Many datasets available
– Adopted Big Buck Bunny & DASHed it
• Players
– bitdash
– Proprietary solutions (smooth, HLS, HDS)
– YouTube, dash.js, DASH-JS
– …and compare it with ten different adaptation algorithms
• Objective evaluation
– Common test setup using network emulation & bandwidth shaping
– Predefined bandwidth trajectory (or real network traces)
• Subjective evaluation
– Lab [ITU-T B.500 / P.910] vs.
crowdsourcing with special
platforms or social networks
July 2016 ICME 2016 Tutorial, C. Timmerer 16
89. Crowdsourced QoE Evaluation
• Quality of Experience …
– Mean Opinion Score [0..100]
– [other objective metrics:
start-up time, throughput, stalls]
• … Web-based Adaptive HTTP Streaming Clients …
– HTML5+MSE: DASH-JS (dash.itec.aau.at), dash.js (DASH-IF, v1.1.2), YouTube
• … Real-World Environments …
– DASH-JS, dash.js hosted at ITEC/AAU (~ 10Gbit/s)
– YouTube hosted at Google data centers
– Content: Tears of Steel @ 144p (250 kbit/s), 240p (380 kbit/s), 360p (740 kbit/s), 480p (1308 kbit/s), and
720p (2300 kbit/s); segment size: 2s
– Users access content over the open Internet (i.e., real-world environment)
• … Crowdsourcing
– Campaign at Microworker platform (others also possible: Mechanical Turk, social networks) limited to
Europe, USA/Canada, India
– Screening Techniques: Browser fingerprinting, stimulus presentation time, QoE ratings and pre-
questionnaire
July 2016 ICME 2016 Tutorial, C. Timmerer 17
B. Rainer, C. Timmerer, “Quality ofExperience ofWeb-based Adaptive HTTP Streaming Clients in Real-World Environments using Crowdsourcing”,
Proceedings of International Workshop on VideoNext: Design,Quality andDeployment of Adaptive Video Streaming, Sydney, Australia,Dec. 2014.
92. Results Summary
• DASH-JS
– High start-up time
– Low number of stalls
– Good throughput, QoE
• dash.js
– Low start-up time
– High # stalls
– Low throughput
– Low QoE
• YouTube
– Low start-up time
– Low number of stalls
– Best throughput, QoE
July 2016 ICME 2016 Tutorial, C. Timmerer 20
96. DASH-JS vs. bitdash
July 2016 ICME 2016 Tutorial, C. Timmerer 24
C. Timmerer, D. Weinberger, C. Mueller, and S. Lederer, “Ultra-High-Definition-Quality of Experience with MPEG-DASH”,
Proceedings of the Broadcast EngineeringConference (BEC), NAB2015, Las Vegas, NV, USA, April 2015.
99. Conclusions (1)
• MPEG-DASH defines formats only
– Media Presentation Description (MPD)
– Segment format: mp4, ts
• MPEG-DASH is not
– System, protocol, presentation, codec, interactivity, DRM, client specification
– Other standards required for a complete ecosystem: e.g., DASH-IF, WAVE,
HMTL5, MSE, EME
• Do we need MPEG-DASH? (for adaptive media streaming)
– Not necessarily: e.g., WebM + VPx + manifest & control end-to-end
– Required to address heterogeneous environments
• Role of standards sometimes overrated but often underestimated
July 2016 ICME 2016 Tutorial, C. Timmerer 27
100. Conclusions (2)
• QoE for DASH-based services (a rule of thumb)
– Startup delay (low [but live vs. on-demand & short vs. long-tail
content])
– Buffer underrun / stalls (zero)
– Quality switches (low) and media throughput (high)
– Energy- and cost-awareness (data plan)
• No general applicable QoE model for DASH
– (Too) many factors influencing / features of QoE for DASH-based
services
– Methodology for reproducible research is in place and well established
– Ample research opportunities
July 2016 ICME 2016 Tutorial, C. Timmerer 28
Main QoE
factors for DASH
Come up with our own QoE factor and design, conduct, analyze a small-scale experiment
103. How to create, delivery, consume?
• Sensory Effect Description Language (SEDL)
– Basic building blocks to describe, e.g., light, wind, fog, vibration, scent
– MPEG-V Part 3, Sensory Information: Effects, GroupOfEffects
– Adopted MPEG-21 DIA tools for adding time information (synchronization)
• Description conforming to SEDL :== Sensory Effect Metadata (SEM)
– Can be associated to any kind of multimedia content (e.g., movies, music, Web
sites, games)
– Support to be included in file (MP4) and transport (M2TS) formats
• Tool support for creating (annotation tools) and consumption (players,
Web plugins) selab.itec.aau.at
• Devices: e.g., amBX (Ambient Experience) system + SDK,
Gameskunk, Scentscape, etc.
July 2016 ICME 2016 Tutorial, C. Timmerer 31
104. How to capture and measure?
• Subjective quality assessments
– Methodology: based on standard methods
– Test content: different genres, manually annotated (cf. QUALINET DB)
• Experiment I
– Aim: Demonstrate sensory effects as a vital tool for enhancing the quality of experience
depending on the actual genre
• Experiment II
– Aim: investigate the relationship of the QoE to various video bit-rates of multimedia contents
annotated with sensory effects.
– Subjective quality gap between video resources annotated with and without sensory effects at
different bit-rates
• [Experiment III] ambient lights & different color calculation settings
• Experiment IV
– Aim: investigate the enhancement of the QoE and how users’ emotions are elicited and
influenced by Web videos annotated with and without sensory effects
July 2016 ICME 2016 Tutorial, C. Timmerer 32
105. Experiment II: Results
July 2016 ICME 2016 Tutorial, C. Timmerer 33
Sequence Babylon A.D. Earth
Duration 35s 21s
Resolution 1280 x 544 1280 x 720
Motion High Low
Nr. of Effects W: 7; V: 9 W: 8; V: 1
Bit-rates Kbit/s PSNR Kbit/s PSNR
Low Quality 2154 38.93 2204 38.11
Medium Quality 3112 41.27 3171 40.65
High Quality 4044 42.95 4116 42.27
Highest Quality 6315 N/A 6701 N/A
Test Sequences
MOS vs. PSNR/bit-rate for Earth.
107. How to judge and explain?
• Experiment VI
– Aim: understand QuaSE
• Biosensor-based QoE evaluation system
July 2016 ICME 2016 Tutorial, C. Timmerer 35
J. Donley, C. Ritz, M. Shujau, "Analysing the Quality of Experience
of Multisensory Media from Measurements of Physiological
Responses,” QoMEX2014, Singapore, Sep. 2014.
108. How to judge and explain?
• Experiment VII
– Aim: understand QuaSE
• EEG Correlates of Pleasant and Unpleasant Odor Perception
July 2016 ICME 2016 Tutorial, C. Timmerer 36
E. Kroupi, A. Yazdani, J.-M. Vesin, T. Ebrahimi, "EEG Correlates of Pleasant and
Unpleasant Odor Perception," ACM TOMM, vol. 11, no. 1s, Sep. 2014.
109. How to judge and explain?
• Experiment VIII
– Aim: understand QuaSE
• Multiple-Scent Enhanced Multimedia Synchronization
July 2016 ICME 2016 Tutorial, C. Timmerer 37
N. Murray, B. Lee, Y. Qiao, and G.-M. Muntean, "Multiple-Scent Enhanced Multimedia Synchronization," ACM TOMM, vol. 11, no. 1s, Sep. 2014.
General temporal boundaries:
-10s to +15s are “in-sync”, skew values beyond are “out-of-sync”
111. Open Issues / Challenges
• QoE assessment is a delicate mixture of ingredients and choices
– Test & lab environment
– Test content
– Test methodology
– Data analysis
• (Semi-)Automatic content creation/annotation
• Towards large scale deployment
– Lessons learnt from 3D (disaster)
– 4D, 5D, xD – adding another dimension does not guarantee success
• Holistic approach not feasible
– Need for much more specialized QuaSE models
• QUALINET Task Force: "Immersive Media Experiences (IMEx)”
– https://www3.informatik.uni-wuerzburg.de/qoewiki/qualinet:imex
July 2016 ICME 2016 Tutorial, C. Timmerer 39
114. Touradj Ebrahimi: About Me …
• Professor of multimedia signal processing at EPFL
• Active in image/video compression, media interpretation
(segmentation, annotation, search, retrieval, quality assessment,
brain computer interface, affective computing, etc.) and media
security (privacy protection, copyright protection, media integrity
verification, etc.)
• Member of MPEG standardization committee since 1992 and active
in many of its video standardization activities: MPEG-4, H.264/AVC,
MVC, H.265/HEVC, MV-HEVC, 3D-HEVC, SCC, HDR extensions.
• Member of JPEG standardization committee since 1994 and active in
many of its image standardization activities: JPEG 2000, JPSearch,
JPEG XR, JPEG AIC, JPEG XT, JPEG XS, JPEG PLENO.
• Member of the Steering Committee of QoMEX and chair of its first
edition in 2009
• Convener of JPEG Standardization Committee since 2014
• Chair of COST Action IC1003 Qualinet
• First coined in February 2001 the term Quality of Experience (QoE) as
a user-centric alternative to Quality of Service (QoS)
145. Devices and sensors are Queens!
• (Easily available) Wearable
devices and sensors are
needed to generate the
data:
– Affordable components and
sensors to be purchased by
interested individuals
– Reliable
– Easy to configure and
calibrate
– User friendly
158. Swallowing detection
Use EMG to detect swallowing and enable/disable camera
• The sound of mastication (food crushing) has relation to
physical properties of the food, but little relevance to
energy content
Chewing sensors
O. Amft, M. Stäger, and G. Tröster, “Analysis of chewing sounds for dietary monitoring,” UbiComp 2005, pp. 56–72, 2005.
S. Päßler, M. Wolff, and W.-J. Fischer, “Food intake monitoring: an acoustical approach to automated food intake activity detection
and classification of consumed food,” Physiol. Meas., vol. 33, no. 6, pp. 1073–1093, 2012.
• The sound of mastication (food crushing) has relation to
physical properties of the food, but little relevance to
energy content
Chewing sensors
O. Amft, M. Stäger, and G. Tröster, “Analysis of chewing sounds for dietary monitoring,” UbiComp 2005, pp. 56–72, 2005.
S. Päßler, M. Wolff, and W.-J. Fischer, “Food intake monitoring: an acoustical approach to automated food intake activity detection
and classification of consumed food,” Physiol. Meas., vol. 33, no. 6, pp. 1073–1093, 2012.
162. Take-Home Messages
• QoL is the natural step beyond QoE !
• QoL not a new concept but it can take advantage of
modern technologies.
• A federating project is needed in order to create the
necessary critical mass (especially in data).
• Multimedia Dietary Assessment is a compelling use
case.