SlideShare a Scribd company logo
1 of 26
Download to read offline
Akademia G´orniczo-Hutnicza
im. Stanislawa Staszica w Krakowie
AGH University of Science
and Technology
Video Quality Assessment in
Recognition Tasks
Kamil Kawa1 Mikolaj Leszczuk1 Atanas Boev2
1AGH University of Science and Technology, PL-30059 Krak´ow, Poland
vq@kt.agh.edu.pl
http://vq.kt.agh.edu.pl
2Huawei Technologies Duesseldorf GmbH, 40549 Duesseldorf, Germany
atanas.boev@huawei.com
Kamil Kawa, Mikolaj Leszczuk, Atanas Boev Video Quality Assessment in Recognition Tasks 1 / 23
www.agh.edu.pl
Outline
1 Problem Introduction
Kamil Kawa, Mikolaj Leszczuk, Atanas Boev Video Quality Assessment in Recognition Tasks 2 / 23
www.agh.edu.pl
Outline
1 Problem Introduction
2 Assessments Environment
General Viewing Conditions for Subjective Assessments in Laboratory
Environment
General Viewing Conditions for Subjective Assessments in Home
Environment
Kamil Kawa, Mikolaj Leszczuk, Atanas Boev Video Quality Assessment in Recognition Tasks 2 / 23
www.agh.edu.pl
Outline
1 Problem Introduction
2 Assessments Environment
General Viewing Conditions for Subjective Assessments in Laboratory
Environment
General Viewing Conditions for Subjective Assessments in Home
Environment
3 Methods
Single Choice Method
Multiple Choice Method
Timed Task Method
Scenes
Kamil Kawa, Mikolaj Leszczuk, Atanas Boev Video Quality Assessment in Recognition Tasks 2 / 23
www.agh.edu.pl
Outline
1 Problem Introduction
2 Assessments Environment
General Viewing Conditions for Subjective Assessments in Laboratory
Environment
General Viewing Conditions for Subjective Assessments in Home
Environment
3 Methods
Single Choice Method
Multiple Choice Method
Timed Task Method
Scenes
4 Conclusion
Kamil Kawa, Mikolaj Leszczuk, Atanas Boev Video Quality Assessment in Recognition Tasks 2 / 23
www.agh.edu.pl Problem Introduction
Problem Introduction
Nowadays, many metrics for overall Quality of Experience (QoE),
successfully used in video processing systems for video quality
evaluation
Both:
Full-Reference ones, like Peak Signal–to–Noise Ratio – PSNR or
Structural Similarity – SSIM
Non-Reference ones, like video quality indicators
However – not appropriate for recognition tasks analytic (Target
Recognition Video, TRV)
Kamil Kawa, Mikolaj Leszczuk, Atanas Boev Video Quality Assessment in Recognition Tasks 3 / 23
www.agh.edu.pl Problem Introduction
Problem Introduction
“Target” – object on video that tester needs to identify, e.g.:
Face
Object,
Number
TRV – video used as tool checking ability to recognise specific targets
of interests in video stream
TRV applicable in various services such as:
Surveillance
Licence place identification
Human identification
Telemedicine
According to, one can divide the target into three categories:
1 Human identification (including facial recognition)
2 Object identification
3 Alphanumeric identification
Kamil Kawa, Mikolaj Leszczuk, Atanas Boev Video Quality Assessment in Recognition Tasks 4 / 23
www.agh.edu.pl Problem Introduction
Problem Introduction
Given use of TRV, qualitative tests:
Not focusing on the subject’s satisfaction with video sequence quality
Measure how subject using TRV to accomplish certain tasks
Example purposes of this:
Video surveillance – recognition of vehicle license plate numbers
Telemedicine/remote diagnostics – correct diagnosis
Fire safety – fire detection
Rear backup cameras – parking the car
Games – spotting and correctly reacting to virtual enemy
Video newscasts and reports editing – video summarization
Kamil Kawa, Mikolaj Leszczuk, Atanas Boev Video Quality Assessment in Recognition Tasks 5 / 23
www.agh.edu.pl Problem Introduction
Problem Introduction
Traditional approach to video quality assessment mostly focusing on
Quality of Service (QoS) techniques
However – obsolete method now
To prepare more accurate assessment of video quality, one to take into
account perception of user
Based on limitation of QoS for video applications, QoE describing
performance of whole, end-to-end video delivery system from user’s
point of view
Several essential factors affecting perceived video QoE:
Quality degradation during the content production phase
Artefact introduced by lossy compression
Network transmission errors
Application and display device-specific parameters
End-user’s preferences and perception model
Kamil Kawa, Mikolaj Leszczuk, Atanas Boev Video Quality Assessment in Recognition Tasks 6 / 23
www.agh.edu.pl Problem Introduction
Problem Introduction
Moreover, methods and metrics, to fulfil following expectations:
In-service applicable
Non-reference quality assessment
High performance for diverse video content
Coverage of all mentioned factors contributing to overall QoE
Mapping between measured parameters (QoS, artefacts level) and QoE
Kamil Kawa, Mikolaj Leszczuk, Atanas Boev Video Quality Assessment in Recognition Tasks 7 / 23
www.agh.edu.pl Problem Introduction
Problem Introduction
Many parameters impacting ability to achieve recognition task, but
selecting five of them as most important ones:
1 Usage time frame – specifying if one in need to analyse video in
real-time or to be stored and analysed later
2 Discrimination level – specifying fine level of detail sought from video
3 Target size – specifying whether predicted region of interest in video
occupies relatively small or large percentage of video
4 Lightning level – specifying anticipated lighting level of scene
5 Level of motion – specifying anticipated motion level in video scene
Kamil Kawa, Mikolaj Leszczuk, Atanas Boev Video Quality Assessment in Recognition Tasks 8 / 23
www.agh.edu.pl Assessments Environment
Assessments Environment
General viewing condition for subjective assessments to be met.
Conditions divided into:
Home environment
Laboratory environment
Kamil Kawa, Mikolaj Leszczuk, Atanas Boev Video Quality Assessment in Recognition Tasks 9 / 23
www.agh.edu.pl Assessments Environment General Viewing Conditions for Subjective Assessments in Laboratory Environment
General Viewing Conditions for Assessments in Laboratory Environment
The assessors’ viewing conditions should be arranged as follows:
Table: Viewing condition for subjective assessments in laboratory environment.
Ratio of luminance of inactive screen to peak luminance: <=0.02
Ratio of the luminance of the screen, when displaying only black
level in a completely dark room, to that corresponding to peak white:
≈ 0.01
Maximum observation angle relative to the normal 30
Ratio of luminance of background behind picture monitor
to peak luminance of picture
≈ 0.15
Chromaticity of background D65
Other room illumination low
Kamil Kawa, Mikolaj Leszczuk, Atanas Boev Video Quality Assessment in Recognition Tasks 10 / 23
www.agh.edu.pl Assessments Environment General Viewing Conditions for Subjective Assessments in Home Environment
General Viewing Conditions for Assessments in Home Environment
Viewing distance and screen size to be selected in order to satisfy PVD
(Preferred Viewing Distance).
Table: Viewing condition for subjective assessments in home environment.
Inactive screen vs. peak luminance <=0.02
Maximum relative vs. normal observation angle 30◦
Screen size for a 4/3 format ratio Screen size to satisfy PVD rules
Screen size for a 16/9 format ratio Screen size to satisfy PVD rules
Monitor processing without digital processing
Peak luminance 200 cd/m2
Environmental illuminance on the screen 200lux
Kamil Kawa, Mikolaj Leszczuk, Atanas Boev Video Quality Assessment in Recognition Tasks 11 / 23
www.agh.edu.pl Assessments Environment General Viewing Conditions for Subjective Assessments in Home Environment
General Viewing Conditions for Assessments in Home Environment
PVD in function of screen size shown in Table
Information in table and function recommending PVD related screen
size that should be used
Table: Information on PVD and related screen sizes.
Screen diagonal
(in)
Screen height
(H)
PVD
4/3 ratio 16/9 ratio (m) (H)
12 15 0.18 9
15 18 0.23 8
20 24 0.30 7
29 36 0.45 6
60 73 0.91 5
>100 >120 >1.53 3-5
Kamil Kawa, Mikolaj Leszczuk, Atanas Boev Video Quality Assessment in Recognition Tasks 12 / 23
www.agh.edu.pl Assessments Environment General Viewing Conditions for Subjective Assessments in Home Environment
General Viewing Conditions for Assessments in Home Environment
Figure: PVD for moving images.
Kamil Kawa, Mikolaj Leszczuk, Atanas Boev Video Quality Assessment in Recognition Tasks 13 / 23
www.agh.edu.pl Methods
Methods
ITU-T P.912 Recommendation introducing lot of useful methods for
recognition tasks
This Recommendation defining subjective assessment methods for
evaluating quality of one-way video used for target recognition tasks
Kamil Kawa, Mikolaj Leszczuk, Atanas Boev Video Quality Assessment in Recognition Tasks 14 / 23
www.agh.edu.pl Methods Single Choice Method
Single Choice Method
Method to be used when single,
unambiguous answer to
identification question
Technique utilisable for
alphanumeric character
recognition scenarios
Experimenter asking tester which
letter(s), or number(s) appearing
in specific area of video
Answer to be evaluated only as
binary one:
Correct
Incorrect
According to that, Yes or No
test – also acceptable
Kamil Kawa, Mikolaj Leszczuk, Atanas Boev Video Quality Assessment in Recognition Tasks 15 / 23
www.agh.edu.pl Methods Single Choice Method
Single Choice Method
For example, one asking viewer
if certain object present in scene
In such a method – essential to
ensure availability of easy to
understand answers
Care also to be taken to avoid
terminology differences
Use of “unsure” answer allowed,
but not recommended
Reason for that: over-usage of
this option by testers, leading to
contamination of experiment
result
Kamil Kawa, Mikolaj Leszczuk, Atanas Boev Video Quality Assessment in Recognition Tasks 16 / 23
www.agh.edu.pl Methods Multiple Choice Method
Multiple Choice Method
This method especially
appropriate for all discrimination
class levels (introduced in ITU-T
P.912) and target categories
For this method, experimenter
showing:
Video (this slide)
List of possible answers (next
slide)
After presenting video, viewers
to choose label being closest to
what recognised on clip
Kamil Kawa, Mikolaj Leszczuk, Atanas Boev Video Quality Assessment in Recognition Tasks 17 / 23
www.agh.edu.pl Methods Multiple Choice Method
Multiple Choice Method
Use of fixed multiple choices
eliminating any possible
misunderstanding possibly
arising from open questions
Due to that, more accurate
measurements possible
Number of choices offered to
tester depending on number of
alternative scenes presented
As in previous method, one to
take special care when ”unsure”
is one of listed choices
Kamil Kawa, Mikolaj Leszczuk, Atanas Boev Video Quality Assessment in Recognition Tasks 18 / 23
www.agh.edu.pl Methods Timed Task Method
Timed Task Method
Viewer to be asked to watch for particular action or object that viewer
about to recognise in video clip
When tester perceiving target occurrence, pushing button
In timed task, experimenter to determine whether time falling within
acceptable time-frame for decision making
This time-frames applicable for example in video scenarios:
“A tester is responding to a violent situation and needs to identify
whether crowd members have real weapons.”
“A person is chasing a vehicle and needs to read the license plate.”
Kamil Kawa, Mikolaj Leszczuk, Atanas Boev Video Quality Assessment in Recognition Tasks 19 / 23
www.agh.edu.pl Methods Scenes
Scenes
TRV used to perform a
recognition task
Scenes to contain targets
consistent with the application
under study
Measurement of test mostly
focused on subject’s ability to
identify objects and actions
Problem: viewer possibly
memorising scene content and
using other visual clues to
remember identity of target
Therefore, set of scenes
containing multiple versions to
replace particular scene
Kamil Kawa, Mikolaj Leszczuk, Atanas Boev Video Quality Assessment in Recognition Tasks 20 / 23
www.agh.edu.pl Methods Scenes
Scenes
It is best way to control
differences between versions
Example scenario:
“A person walks across the
field of view carrying objects.”
Set of videos to consist of
multiple shots using different
object and different people
Number of scenarios in set to be
large enough to reduce likelihood
of scene memorisation
Of course, experts to determine
content of each scene
Kamil Kawa, Mikolaj Leszczuk, Atanas Boev Video Quality Assessment in Recognition Tasks 21 / 23
www.agh.edu.pl Methods Scenes
Scenes
Difference to be, for example,
object carried by person on video
Experts to identify critical tasks
or parameters of scenes
One to base set of
multiple-choice answer and
experiment design on these
parameters
One to create scene in way that,
object of interest appearing in
video at resolution realistically
expected in practice
Kamil Kawa, Mikolaj Leszczuk, Atanas Boev Video Quality Assessment in Recognition Tasks 22 / 23
www.agh.edu.pl Conclusion
Conclusion
More general review of possible methods that could be used to video
quality assessment presented in paper
Based on limitation of QoS for video applications, QoE describing
performance of whole end-to-end video delivery system from user’s
point of view
Key point for every test session: general viewing conditions for
subjective assessments in laboratory and home environment
Various methods applicable to recognition task defined in paper
Each method containing list with brief description of used methods in
scientific experiments
Kamil Kawa, Mikolaj Leszczuk, Atanas Boev Video Quality Assessment in Recognition Tasks 23 / 23

More Related Content

Similar to Survey on the State-Of-The-Art Methods for Objective Video Quality Assessment in Recognition Tasks

Video Quality Measurements
Video Quality MeasurementsVideo Quality Measurements
Video Quality MeasurementsYoss Cohen
 
Pov09 Online Presentation
Pov09 Online PresentationPov09 Online Presentation
Pov09 Online Presentationkatrinas
 
Video Quality Measurement based on Network Traffic
Video Quality Measurement based on Network TrafficVideo Quality Measurement based on Network Traffic
Video Quality Measurement based on Network TrafficAmir Hossein Mandegar
 
Quality of Multimedia Experience: Past, Present and Future
Quality of Multimedia Experience: Past, Present and FutureQuality of Multimedia Experience: Past, Present and Future
Quality of Multimedia Experience: Past, Present and FutureTouradj Ebrahimi
 
MPEG Visual Quality Assessment: Tasks and Perspectives
MPEG Visual Quality Assessment: Tasks and PerspectivesMPEG Visual Quality Assessment: Tasks and Perspectives
MPEG Visual Quality Assessment: Tasks and PerspectivesAlpen-Adria-Universität
 
Full reference video quality assessment
Full reference video quality assessmentFull reference video quality assessment
Full reference video quality assessmentHoàng Sơn
 
SUBJECTIVE QUALITY EVALUATION OF H.264 AND H.265 ENCODED VIDEO SEQUENCES STRE...
SUBJECTIVE QUALITY EVALUATION OF H.264 AND H.265 ENCODED VIDEO SEQUENCES STRE...SUBJECTIVE QUALITY EVALUATION OF H.264 AND H.265 ENCODED VIDEO SEQUENCES STRE...
SUBJECTIVE QUALITY EVALUATION OF H.264 AND H.265 ENCODED VIDEO SEQUENCES STRE...ijma
 
Subjective Quality Evaluation of H.264 and H.265 Encoded Video Sequences Stre...
Subjective Quality Evaluation of H.264 and H.265 Encoded Video Sequences Stre...Subjective Quality Evaluation of H.264 and H.265 Encoded Video Sequences Stre...
Subjective Quality Evaluation of H.264 and H.265 Encoded Video Sequences Stre...ijma
 
Video Testing Best Practices: How to Guarantee High-Quality Video for your Cu...
Video Testing Best Practices: How to Guarantee High-Quality Video for your Cu...Video Testing Best Practices: How to Guarantee High-Quality Video for your Cu...
Video Testing Best Practices: How to Guarantee High-Quality Video for your Cu...Perfecto by Perforce
 
Ultra-High-Definition Quality of Experience with MPEG-DASH
Ultra-High-Definition Quality of Experience with MPEG-DASHUltra-High-Definition Quality of Experience with MPEG-DASH
Ultra-High-Definition Quality of Experience with MPEG-DASHBitmovin Inc
 
A survey on Measurement of Objective Video Quality in Social Cloud using Mach...
A survey on Measurement of Objective Video Quality in Social Cloud using Mach...A survey on Measurement of Objective Video Quality in Social Cloud using Mach...
A survey on Measurement of Objective Video Quality in Social Cloud using Mach...IRJET Journal
 
Video Captioning at TRECVID 2022
Video Captioning at TRECVID 2022Video Captioning at TRECVID 2022
Video Captioning at TRECVID 2022George Awad
 
Iec 62676 5 standardized spezification of image quality
Iec 62676 5 standardized spezification of image qualityIec 62676 5 standardized spezification of image quality
Iec 62676 5 standardized spezification of image qualityHenry Rutjes
 
FutureComm 2010: Video Quality Analysis and Measurement
FutureComm 2010: Video Quality Analysis and MeasurementFutureComm 2010: Video Quality Analysis and Measurement
FutureComm 2010: Video Quality Analysis and MeasurementRADVISION Ltd.
 
Perfecting a user experience research with eye tracking lior yair_orenshtang2
Perfecting a user experience research with eye tracking lior yair_orenshtang2Perfecting a user experience research with eye tracking lior yair_orenshtang2
Perfecting a user experience research with eye tracking lior yair_orenshtang2Lior Yair
 
Machine Vision On Embedded Platform
Machine Vision On Embedded Platform Machine Vision On Embedded Platform
Machine Vision On Embedded Platform Omkar Rane
 
Machine vision Application
Machine vision ApplicationMachine vision Application
Machine vision ApplicationAbhishek Sainkar
 
Voice and Video over IP Communications: Assessing and Improving User Experience
Voice and Video over IP Communications: Assessing and Improving User ExperienceVoice and Video over IP Communications: Assessing and Improving User Experience
Voice and Video over IP Communications: Assessing and Improving User ExperienceRADVISION Ltd.
 

Similar to Survey on the State-Of-The-Art Methods for Objective Video Quality Assessment in Recognition Tasks (20)

Video Quality Measurements
Video Quality MeasurementsVideo Quality Measurements
Video Quality Measurements
 
Pov09 Online Presentation
Pov09 Online PresentationPov09 Online Presentation
Pov09 Online Presentation
 
Video Quality Measurement based on Network Traffic
Video Quality Measurement based on Network TrafficVideo Quality Measurement based on Network Traffic
Video Quality Measurement based on Network Traffic
 
Quality of Multimedia Experience: Past, Present and Future
Quality of Multimedia Experience: Past, Present and FutureQuality of Multimedia Experience: Past, Present and Future
Quality of Multimedia Experience: Past, Present and Future
 
MPEG Visual Quality Assessment: Tasks and Perspectives
MPEG Visual Quality Assessment: Tasks and PerspectivesMPEG Visual Quality Assessment: Tasks and Perspectives
MPEG Visual Quality Assessment: Tasks and Perspectives
 
Full reference video quality assessment
Full reference video quality assessmentFull reference video quality assessment
Full reference video quality assessment
 
01 shewbridge
01 shewbridge01 shewbridge
01 shewbridge
 
SUBJECTIVE QUALITY EVALUATION OF H.264 AND H.265 ENCODED VIDEO SEQUENCES STRE...
SUBJECTIVE QUALITY EVALUATION OF H.264 AND H.265 ENCODED VIDEO SEQUENCES STRE...SUBJECTIVE QUALITY EVALUATION OF H.264 AND H.265 ENCODED VIDEO SEQUENCES STRE...
SUBJECTIVE QUALITY EVALUATION OF H.264 AND H.265 ENCODED VIDEO SEQUENCES STRE...
 
Subjective Quality Evaluation of H.264 and H.265 Encoded Video Sequences Stre...
Subjective Quality Evaluation of H.264 and H.265 Encoded Video Sequences Stre...Subjective Quality Evaluation of H.264 and H.265 Encoded Video Sequences Stre...
Subjective Quality Evaluation of H.264 and H.265 Encoded Video Sequences Stre...
 
Video Testing Best Practices: How to Guarantee High-Quality Video for your Cu...
Video Testing Best Practices: How to Guarantee High-Quality Video for your Cu...Video Testing Best Practices: How to Guarantee High-Quality Video for your Cu...
Video Testing Best Practices: How to Guarantee High-Quality Video for your Cu...
 
Ultra-High-Definition Quality of Experience with MPEG-DASH
Ultra-High-Definition Quality of Experience with MPEG-DASHUltra-High-Definition Quality of Experience with MPEG-DASH
Ultra-High-Definition Quality of Experience with MPEG-DASH
 
A survey on Measurement of Objective Video Quality in Social Cloud using Mach...
A survey on Measurement of Objective Video Quality in Social Cloud using Mach...A survey on Measurement of Objective Video Quality in Social Cloud using Mach...
A survey on Measurement of Objective Video Quality in Social Cloud using Mach...
 
Video Captioning at TRECVID 2022
Video Captioning at TRECVID 2022Video Captioning at TRECVID 2022
Video Captioning at TRECVID 2022
 
Iec 62676 5 standardized spezification of image quality
Iec 62676 5 standardized spezification of image qualityIec 62676 5 standardized spezification of image quality
Iec 62676 5 standardized spezification of image quality
 
FutureComm 2010: Video Quality Analysis and Measurement
FutureComm 2010: Video Quality Analysis and MeasurementFutureComm 2010: Video Quality Analysis and Measurement
FutureComm 2010: Video Quality Analysis and Measurement
 
Perfecting a user experience research with eye tracking lior yair_orenshtang2
Perfecting a user experience research with eye tracking lior yair_orenshtang2Perfecting a user experience research with eye tracking lior yair_orenshtang2
Perfecting a user experience research with eye tracking lior yair_orenshtang2
 
Machine Vision On Embedded Platform
Machine Vision On Embedded Platform Machine Vision On Embedded Platform
Machine Vision On Embedded Platform
 
Machine vision Application
Machine vision ApplicationMachine vision Application
Machine vision Application
 
Voice and Video over IP Communications: Assessing and Improving User Experience
Voice and Video over IP Communications: Assessing and Improving User ExperienceVoice and Video over IP Communications: Assessing and Improving User Experience
Voice and Video over IP Communications: Assessing and Improving User Experience
 
Presentación Tesis 08022016
Presentación Tesis 08022016Presentación Tesis 08022016
Presentación Tesis 08022016
 

More from Mikolaj Leszczuk

Evaluation of Video Summarization
Evaluation of Video SummarizationEvaluation of Video Summarization
Evaluation of Video SummarizationMikolaj Leszczuk
 
Special Session on: Quality Assessment for Computer Vision and Immersive Medi...
Special Session on:Quality Assessment for Computer Vision and Immersive Medi...Special Session on:Quality Assessment for Computer Vision and Immersive Medi...
Special Session on: Quality Assessment for Computer Vision and Immersive Medi...Mikolaj Leszczuk
 
Self-Improving Sustainable Intelligent Transport System (ITS) Using Video Con...
Self-Improving Sustainable Intelligent Transport System (ITS) Using Video Con...Self-Improving Sustainable Intelligent Transport System (ITS) Using Video Con...
Self-Improving Sustainable Intelligent Transport System (ITS) Using Video Con...Mikolaj Leszczuk
 
#Paris Meeting 2018 - Presentation of @chist_era_AMIS
#Paris Meeting 2018 - Presentation of @chist_era_AMIS#Paris Meeting 2018 - Presentation of @chist_era_AMIS
#Paris Meeting 2018 - Presentation of @chist_era_AMISMikolaj Leszczuk
 
Spotkanie w VIII Prywatnym Akademickim Liceum Ogólnokształcącym
Spotkanie w VIII Prywatnym Akademickim Liceum OgólnokształcącymSpotkanie w VIII Prywatnym Akademickim Liceum Ogólnokształcącym
Spotkanie w VIII Prywatnym Akademickim Liceum OgólnokształcącymMikolaj Leszczuk
 
Prace naukowe prowadzone w Katedrze Telekomunikacji @AGH_Krakow
Prace naukowe prowadzone w Katedrze Telekomunikacji @AGH_KrakowPrace naukowe prowadzone w Katedrze Telekomunikacji @AGH_Krakow
Prace naukowe prowadzone w Katedrze Telekomunikacji @AGH_KrakowMikolaj Leszczuk
 
Infrastructure for High-Attendance, Simple Psychophysical Experiments
Infrastructure for High-Attendance, Simple Psychophysical ExperimentsInfrastructure for High-Attendance, Simple Psychophysical Experiments
Infrastructure for High-Attendance, Simple Psychophysical ExperimentsMikolaj Leszczuk
 
J. Imaging: Special Issue on Image Quality
J. Imaging: Special Issue on Image QualityJ. Imaging: Special Issue on Image Quality
J. Imaging: Special Issue on Image QualityMikolaj Leszczuk
 
Video summarization framework for newscasts and reports – work in progress
Video summarization framework for newscasts and reports – work in progressVideo summarization framework for newscasts and reports – work in progress
Video summarization framework for newscasts and reports – work in progressMikolaj Leszczuk
 
Visual Analytics of Smart City Data for Sustainable Quality of Life of Citizens
Visual Analytics of Smart City Data for Sustainable Quality of Life of CitizensVisual Analytics of Smart City Data for Sustainable Quality of Life of Citizens
Visual Analytics of Smart City Data for Sustainable Quality of Life of CitizensMikolaj Leszczuk
 
Automatic Extraction of Machine Tags in Flickr Service
Automatic Extraction of Machine Tags in Flickr ServiceAutomatic Extraction of Machine Tags in Flickr Service
Automatic Extraction of Machine Tags in Flickr ServiceMikolaj Leszczuk
 
Results on video summarization
Results on video summarizationResults on video summarization
Results on video summarizationMikolaj Leszczuk
 
Człowiek, ósma warstwa modelu ISO/OSI, jako element ekosystemu teleinformaty...
Człowiek, ósma warstwa modelu ISO/OSI, jako element ekosystemu teleinformaty...Człowiek, ósma warstwa modelu ISO/OSI, jako element ekosystemu teleinformaty...
Człowiek, ósma warstwa modelu ISO/OSI, jako element ekosystemu teleinformaty...Mikolaj Leszczuk
 
Badanie i implementacja aspektu QoE (ang. Quality of Experience) w aplikacjac...
Badanie i implementacja aspektu QoE (ang. Quality of Experience) w aplikacjac...Badanie i implementacja aspektu QoE (ang. Quality of Experience) w aplikacjac...
Badanie i implementacja aspektu QoE (ang. Quality of Experience) w aplikacjac...Mikolaj Leszczuk
 
Platforma do automatycznej, obiektywnej oceny jakości usług transmisji wideo
Platforma do automatycznej, obiektywnej oceny jakości usług transmisji wideoPlatforma do automatycznej, obiektywnej oceny jakości usług transmisji wideo
Platforma do automatycznej, obiektywnej oceny jakości usług transmisji wideoMikolaj Leszczuk
 
Modelling of Quality of Experience in No-Reference (NR) Model
Modelling of Quality of Experience in No-Reference (NR) ModelModelling of Quality of Experience in No-Reference (NR) Model
Modelling of Quality of Experience in No-Reference (NR) ModelMikolaj Leszczuk
 
Definition of Requirements for Accessing Multilingual Information and Opinions
Definition of Requirements for Accessing Multilingual Information and OpinionsDefinition of Requirements for Accessing Multilingual Information and Opinions
Definition of Requirements for Accessing Multilingual Information and OpinionsMikolaj Leszczuk
 
Aplikacja mobilna do rozpoznawania numerów linii komunikacji miejskiej
Aplikacja mobilna do rozpoznawania numerów linii komunikacji miejskiejAplikacja mobilna do rozpoznawania numerów linii komunikacji miejskiej
Aplikacja mobilna do rozpoznawania numerów linii komunikacji miejskiejMikolaj Leszczuk
 

More from Mikolaj Leszczuk (20)

#VQEG #QUADRIVIA 2020
#VQEG #QUADRIVIA 2020#VQEG #QUADRIVIA 2020
#VQEG #QUADRIVIA 2020
 
Evaluation of Video Summarization
Evaluation of Video SummarizationEvaluation of Video Summarization
Evaluation of Video Summarization
 
Special Session on: Quality Assessment for Computer Vision and Immersive Medi...
Special Session on:Quality Assessment for Computer Vision and Immersive Medi...Special Session on:Quality Assessment for Computer Vision and Immersive Medi...
Special Session on: Quality Assessment for Computer Vision and Immersive Medi...
 
Self-Improving Sustainable Intelligent Transport System (ITS) Using Video Con...
Self-Improving Sustainable Intelligent Transport System (ITS) Using Video Con...Self-Improving Sustainable Intelligent Transport System (ITS) Using Video Con...
Self-Improving Sustainable Intelligent Transport System (ITS) Using Video Con...
 
#Paris Meeting 2018 - Presentation of @chist_era_AMIS
#Paris Meeting 2018 - Presentation of @chist_era_AMIS#Paris Meeting 2018 - Presentation of @chist_era_AMIS
#Paris Meeting 2018 - Presentation of @chist_era_AMIS
 
Spotkanie w VIII Prywatnym Akademickim Liceum Ogólnokształcącym
Spotkanie w VIII Prywatnym Akademickim Liceum OgólnokształcącymSpotkanie w VIII Prywatnym Akademickim Liceum Ogólnokształcącym
Spotkanie w VIII Prywatnym Akademickim Liceum Ogólnokształcącym
 
QoE Research
QoE ResearchQoE Research
QoE Research
 
Prace naukowe prowadzone w Katedrze Telekomunikacji @AGH_Krakow
Prace naukowe prowadzone w Katedrze Telekomunikacji @AGH_KrakowPrace naukowe prowadzone w Katedrze Telekomunikacji @AGH_Krakow
Prace naukowe prowadzone w Katedrze Telekomunikacji @AGH_Krakow
 
Infrastructure for High-Attendance, Simple Psychophysical Experiments
Infrastructure for High-Attendance, Simple Psychophysical ExperimentsInfrastructure for High-Attendance, Simple Psychophysical Experiments
Infrastructure for High-Attendance, Simple Psychophysical Experiments
 
J. Imaging: Special Issue on Image Quality
J. Imaging: Special Issue on Image QualityJ. Imaging: Special Issue on Image Quality
J. Imaging: Special Issue on Image Quality
 
Video summarization framework for newscasts and reports – work in progress
Video summarization framework for newscasts and reports – work in progressVideo summarization framework for newscasts and reports – work in progress
Video summarization framework for newscasts and reports – work in progress
 
Visual Analytics of Smart City Data for Sustainable Quality of Life of Citizens
Visual Analytics of Smart City Data for Sustainable Quality of Life of CitizensVisual Analytics of Smart City Data for Sustainable Quality of Life of Citizens
Visual Analytics of Smart City Data for Sustainable Quality of Life of Citizens
 
Automatic Extraction of Machine Tags in Flickr Service
Automatic Extraction of Machine Tags in Flickr ServiceAutomatic Extraction of Machine Tags in Flickr Service
Automatic Extraction of Machine Tags in Flickr Service
 
Results on video summarization
Results on video summarizationResults on video summarization
Results on video summarization
 
Człowiek, ósma warstwa modelu ISO/OSI, jako element ekosystemu teleinformaty...
Człowiek, ósma warstwa modelu ISO/OSI, jako element ekosystemu teleinformaty...Człowiek, ósma warstwa modelu ISO/OSI, jako element ekosystemu teleinformaty...
Człowiek, ósma warstwa modelu ISO/OSI, jako element ekosystemu teleinformaty...
 
Badanie i implementacja aspektu QoE (ang. Quality of Experience) w aplikacjac...
Badanie i implementacja aspektu QoE (ang. Quality of Experience) w aplikacjac...Badanie i implementacja aspektu QoE (ang. Quality of Experience) w aplikacjac...
Badanie i implementacja aspektu QoE (ang. Quality of Experience) w aplikacjac...
 
Platforma do automatycznej, obiektywnej oceny jakości usług transmisji wideo
Platforma do automatycznej, obiektywnej oceny jakości usług transmisji wideoPlatforma do automatycznej, obiektywnej oceny jakości usług transmisji wideo
Platforma do automatycznej, obiektywnej oceny jakości usług transmisji wideo
 
Modelling of Quality of Experience in No-Reference (NR) Model
Modelling of Quality of Experience in No-Reference (NR) ModelModelling of Quality of Experience in No-Reference (NR) Model
Modelling of Quality of Experience in No-Reference (NR) Model
 
Definition of Requirements for Accessing Multilingual Information and Opinions
Definition of Requirements for Accessing Multilingual Information and OpinionsDefinition of Requirements for Accessing Multilingual Information and Opinions
Definition of Requirements for Accessing Multilingual Information and Opinions
 
Aplikacja mobilna do rozpoznawania numerów linii komunikacji miejskiej
Aplikacja mobilna do rozpoznawania numerów linii komunikacji miejskiejAplikacja mobilna do rozpoznawania numerów linii komunikacji miejskiej
Aplikacja mobilna do rozpoznawania numerów linii komunikacji miejskiej
 

Recently uploaded

Recombinant DNA technology( Transgenic plant and animal)
Recombinant DNA technology( Transgenic plant and animal)Recombinant DNA technology( Transgenic plant and animal)
Recombinant DNA technology( Transgenic plant and animal)DHURKADEVIBASKAR
 
Evidences of Evolution General Biology 2
Evidences of Evolution General Biology 2Evidences of Evolution General Biology 2
Evidences of Evolution General Biology 2John Carlo Rollon
 
Pests of safflower_Binomics_Identification_Dr.UPR.pdf
Pests of safflower_Binomics_Identification_Dr.UPR.pdfPests of safflower_Binomics_Identification_Dr.UPR.pdf
Pests of safflower_Binomics_Identification_Dr.UPR.pdfPirithiRaju
 
FREE NURSING BUNDLE FOR NURSES.PDF by na
FREE NURSING BUNDLE FOR NURSES.PDF by naFREE NURSING BUNDLE FOR NURSES.PDF by na
FREE NURSING BUNDLE FOR NURSES.PDF by naJASISJULIANOELYNV
 
TOPIC 8 Temperature and Heat.pdf physics
TOPIC 8 Temperature and Heat.pdf physicsTOPIC 8 Temperature and Heat.pdf physics
TOPIC 8 Temperature and Heat.pdf physicsssuserddc89b
 
STOPPED FLOW METHOD & APPLICATION MURUGAVENI B.pptx
STOPPED FLOW METHOD & APPLICATION MURUGAVENI B.pptxSTOPPED FLOW METHOD & APPLICATION MURUGAVENI B.pptx
STOPPED FLOW METHOD & APPLICATION MURUGAVENI B.pptxMurugaveni B
 
Scheme-of-Work-Science-Stage-4 cambridge science.docx
Scheme-of-Work-Science-Stage-4 cambridge science.docxScheme-of-Work-Science-Stage-4 cambridge science.docx
Scheme-of-Work-Science-Stage-4 cambridge science.docxyaramohamed343013
 
Call Girls in Munirka Delhi 💯Call Us 🔝8264348440🔝
Call Girls in Munirka Delhi 💯Call Us 🔝8264348440🔝Call Girls in Munirka Delhi 💯Call Us 🔝8264348440🔝
Call Girls in Munirka Delhi 💯Call Us 🔝8264348440🔝soniya singh
 
Twin's paradox experiment is a meassurement of the extra dimensions.pptx
Twin's paradox experiment is a meassurement of the extra dimensions.pptxTwin's paradox experiment is a meassurement of the extra dimensions.pptx
Twin's paradox experiment is a meassurement of the extra dimensions.pptxEran Akiva Sinbar
 
‏‏VIRUS - 123455555555555555555555555555555555555555
‏‏VIRUS -  123455555555555555555555555555555555555555‏‏VIRUS -  123455555555555555555555555555555555555555
‏‏VIRUS - 123455555555555555555555555555555555555555kikilily0909
 
Pests of jatropha_Bionomics_identification_Dr.UPR.pdf
Pests of jatropha_Bionomics_identification_Dr.UPR.pdfPests of jatropha_Bionomics_identification_Dr.UPR.pdf
Pests of jatropha_Bionomics_identification_Dr.UPR.pdfPirithiRaju
 
The dark energy paradox leads to a new structure of spacetime.pptx
The dark energy paradox leads to a new structure of spacetime.pptxThe dark energy paradox leads to a new structure of spacetime.pptx
The dark energy paradox leads to a new structure of spacetime.pptxEran Akiva Sinbar
 
BREEDING FOR RESISTANCE TO BIOTIC STRESS.pptx
BREEDING FOR RESISTANCE TO BIOTIC STRESS.pptxBREEDING FOR RESISTANCE TO BIOTIC STRESS.pptx
BREEDING FOR RESISTANCE TO BIOTIC STRESS.pptxPABOLU TEJASREE
 
LIGHT-PHENOMENA-BY-CABUALDIONALDOPANOGANCADIENTE-CONDEZA (1).pptx
LIGHT-PHENOMENA-BY-CABUALDIONALDOPANOGANCADIENTE-CONDEZA (1).pptxLIGHT-PHENOMENA-BY-CABUALDIONALDOPANOGANCADIENTE-CONDEZA (1).pptx
LIGHT-PHENOMENA-BY-CABUALDIONALDOPANOGANCADIENTE-CONDEZA (1).pptxmalonesandreagweneth
 
Environmental Biotechnology Topic:- Microbial Biosensor
Environmental Biotechnology Topic:- Microbial BiosensorEnvironmental Biotechnology Topic:- Microbial Biosensor
Environmental Biotechnology Topic:- Microbial Biosensorsonawaneprad
 
Manassas R - Parkside Middle School 🌎🏫
Manassas R - Parkside Middle School 🌎🏫Manassas R - Parkside Middle School 🌎🏫
Manassas R - Parkside Middle School 🌎🏫qfactory1
 
Microphone- characteristics,carbon microphone, dynamic microphone.pptx
Microphone- characteristics,carbon microphone, dynamic microphone.pptxMicrophone- characteristics,carbon microphone, dynamic microphone.pptx
Microphone- characteristics,carbon microphone, dynamic microphone.pptxpriyankatabhane
 

Recently uploaded (20)

Recombinant DNA technology( Transgenic plant and animal)
Recombinant DNA technology( Transgenic plant and animal)Recombinant DNA technology( Transgenic plant and animal)
Recombinant DNA technology( Transgenic plant and animal)
 
Evidences of Evolution General Biology 2
Evidences of Evolution General Biology 2Evidences of Evolution General Biology 2
Evidences of Evolution General Biology 2
 
Pests of safflower_Binomics_Identification_Dr.UPR.pdf
Pests of safflower_Binomics_Identification_Dr.UPR.pdfPests of safflower_Binomics_Identification_Dr.UPR.pdf
Pests of safflower_Binomics_Identification_Dr.UPR.pdf
 
FREE NURSING BUNDLE FOR NURSES.PDF by na
FREE NURSING BUNDLE FOR NURSES.PDF by naFREE NURSING BUNDLE FOR NURSES.PDF by na
FREE NURSING BUNDLE FOR NURSES.PDF by na
 
TOPIC 8 Temperature and Heat.pdf physics
TOPIC 8 Temperature and Heat.pdf physicsTOPIC 8 Temperature and Heat.pdf physics
TOPIC 8 Temperature and Heat.pdf physics
 
STOPPED FLOW METHOD & APPLICATION MURUGAVENI B.pptx
STOPPED FLOW METHOD & APPLICATION MURUGAVENI B.pptxSTOPPED FLOW METHOD & APPLICATION MURUGAVENI B.pptx
STOPPED FLOW METHOD & APPLICATION MURUGAVENI B.pptx
 
Scheme-of-Work-Science-Stage-4 cambridge science.docx
Scheme-of-Work-Science-Stage-4 cambridge science.docxScheme-of-Work-Science-Stage-4 cambridge science.docx
Scheme-of-Work-Science-Stage-4 cambridge science.docx
 
Call Girls in Munirka Delhi 💯Call Us 🔝8264348440🔝
Call Girls in Munirka Delhi 💯Call Us 🔝8264348440🔝Call Girls in Munirka Delhi 💯Call Us 🔝8264348440🔝
Call Girls in Munirka Delhi 💯Call Us 🔝8264348440🔝
 
Twin's paradox experiment is a meassurement of the extra dimensions.pptx
Twin's paradox experiment is a meassurement of the extra dimensions.pptxTwin's paradox experiment is a meassurement of the extra dimensions.pptx
Twin's paradox experiment is a meassurement of the extra dimensions.pptx
 
‏‏VIRUS - 123455555555555555555555555555555555555555
‏‏VIRUS -  123455555555555555555555555555555555555555‏‏VIRUS -  123455555555555555555555555555555555555555
‏‏VIRUS - 123455555555555555555555555555555555555555
 
Volatile Oils Pharmacognosy And Phytochemistry -I
Volatile Oils Pharmacognosy And Phytochemistry -IVolatile Oils Pharmacognosy And Phytochemistry -I
Volatile Oils Pharmacognosy And Phytochemistry -I
 
Pests of jatropha_Bionomics_identification_Dr.UPR.pdf
Pests of jatropha_Bionomics_identification_Dr.UPR.pdfPests of jatropha_Bionomics_identification_Dr.UPR.pdf
Pests of jatropha_Bionomics_identification_Dr.UPR.pdf
 
The dark energy paradox leads to a new structure of spacetime.pptx
The dark energy paradox leads to a new structure of spacetime.pptxThe dark energy paradox leads to a new structure of spacetime.pptx
The dark energy paradox leads to a new structure of spacetime.pptx
 
BREEDING FOR RESISTANCE TO BIOTIC STRESS.pptx
BREEDING FOR RESISTANCE TO BIOTIC STRESS.pptxBREEDING FOR RESISTANCE TO BIOTIC STRESS.pptx
BREEDING FOR RESISTANCE TO BIOTIC STRESS.pptx
 
LIGHT-PHENOMENA-BY-CABUALDIONALDOPANOGANCADIENTE-CONDEZA (1).pptx
LIGHT-PHENOMENA-BY-CABUALDIONALDOPANOGANCADIENTE-CONDEZA (1).pptxLIGHT-PHENOMENA-BY-CABUALDIONALDOPANOGANCADIENTE-CONDEZA (1).pptx
LIGHT-PHENOMENA-BY-CABUALDIONALDOPANOGANCADIENTE-CONDEZA (1).pptx
 
Environmental Biotechnology Topic:- Microbial Biosensor
Environmental Biotechnology Topic:- Microbial BiosensorEnvironmental Biotechnology Topic:- Microbial Biosensor
Environmental Biotechnology Topic:- Microbial Biosensor
 
Manassas R - Parkside Middle School 🌎🏫
Manassas R - Parkside Middle School 🌎🏫Manassas R - Parkside Middle School 🌎🏫
Manassas R - Parkside Middle School 🌎🏫
 
Microphone- characteristics,carbon microphone, dynamic microphone.pptx
Microphone- characteristics,carbon microphone, dynamic microphone.pptxMicrophone- characteristics,carbon microphone, dynamic microphone.pptx
Microphone- characteristics,carbon microphone, dynamic microphone.pptx
 
Hot Sexy call girls in Moti Nagar,🔝 9953056974 🔝 escort Service
Hot Sexy call girls in  Moti Nagar,🔝 9953056974 🔝 escort ServiceHot Sexy call girls in  Moti Nagar,🔝 9953056974 🔝 escort Service
Hot Sexy call girls in Moti Nagar,🔝 9953056974 🔝 escort Service
 
Engler and Prantl system of classification in plant taxonomy
Engler and Prantl system of classification in plant taxonomyEngler and Prantl system of classification in plant taxonomy
Engler and Prantl system of classification in plant taxonomy
 

Survey on the State-Of-The-Art Methods for Objective Video Quality Assessment in Recognition Tasks

  • 1. Akademia G´orniczo-Hutnicza im. Stanislawa Staszica w Krakowie AGH University of Science and Technology Video Quality Assessment in Recognition Tasks Kamil Kawa1 Mikolaj Leszczuk1 Atanas Boev2 1AGH University of Science and Technology, PL-30059 Krak´ow, Poland vq@kt.agh.edu.pl http://vq.kt.agh.edu.pl 2Huawei Technologies Duesseldorf GmbH, 40549 Duesseldorf, Germany atanas.boev@huawei.com Kamil Kawa, Mikolaj Leszczuk, Atanas Boev Video Quality Assessment in Recognition Tasks 1 / 23
  • 2. www.agh.edu.pl Outline 1 Problem Introduction Kamil Kawa, Mikolaj Leszczuk, Atanas Boev Video Quality Assessment in Recognition Tasks 2 / 23
  • 3. www.agh.edu.pl Outline 1 Problem Introduction 2 Assessments Environment General Viewing Conditions for Subjective Assessments in Laboratory Environment General Viewing Conditions for Subjective Assessments in Home Environment Kamil Kawa, Mikolaj Leszczuk, Atanas Boev Video Quality Assessment in Recognition Tasks 2 / 23
  • 4. www.agh.edu.pl Outline 1 Problem Introduction 2 Assessments Environment General Viewing Conditions for Subjective Assessments in Laboratory Environment General Viewing Conditions for Subjective Assessments in Home Environment 3 Methods Single Choice Method Multiple Choice Method Timed Task Method Scenes Kamil Kawa, Mikolaj Leszczuk, Atanas Boev Video Quality Assessment in Recognition Tasks 2 / 23
  • 5. www.agh.edu.pl Outline 1 Problem Introduction 2 Assessments Environment General Viewing Conditions for Subjective Assessments in Laboratory Environment General Viewing Conditions for Subjective Assessments in Home Environment 3 Methods Single Choice Method Multiple Choice Method Timed Task Method Scenes 4 Conclusion Kamil Kawa, Mikolaj Leszczuk, Atanas Boev Video Quality Assessment in Recognition Tasks 2 / 23
  • 6. www.agh.edu.pl Problem Introduction Problem Introduction Nowadays, many metrics for overall Quality of Experience (QoE), successfully used in video processing systems for video quality evaluation Both: Full-Reference ones, like Peak Signal–to–Noise Ratio – PSNR or Structural Similarity – SSIM Non-Reference ones, like video quality indicators However – not appropriate for recognition tasks analytic (Target Recognition Video, TRV) Kamil Kawa, Mikolaj Leszczuk, Atanas Boev Video Quality Assessment in Recognition Tasks 3 / 23
  • 7. www.agh.edu.pl Problem Introduction Problem Introduction “Target” – object on video that tester needs to identify, e.g.: Face Object, Number TRV – video used as tool checking ability to recognise specific targets of interests in video stream TRV applicable in various services such as: Surveillance Licence place identification Human identification Telemedicine According to, one can divide the target into three categories: 1 Human identification (including facial recognition) 2 Object identification 3 Alphanumeric identification Kamil Kawa, Mikolaj Leszczuk, Atanas Boev Video Quality Assessment in Recognition Tasks 4 / 23
  • 8. www.agh.edu.pl Problem Introduction Problem Introduction Given use of TRV, qualitative tests: Not focusing on the subject’s satisfaction with video sequence quality Measure how subject using TRV to accomplish certain tasks Example purposes of this: Video surveillance – recognition of vehicle license plate numbers Telemedicine/remote diagnostics – correct diagnosis Fire safety – fire detection Rear backup cameras – parking the car Games – spotting and correctly reacting to virtual enemy Video newscasts and reports editing – video summarization Kamil Kawa, Mikolaj Leszczuk, Atanas Boev Video Quality Assessment in Recognition Tasks 5 / 23
  • 9. www.agh.edu.pl Problem Introduction Problem Introduction Traditional approach to video quality assessment mostly focusing on Quality of Service (QoS) techniques However – obsolete method now To prepare more accurate assessment of video quality, one to take into account perception of user Based on limitation of QoS for video applications, QoE describing performance of whole, end-to-end video delivery system from user’s point of view Several essential factors affecting perceived video QoE: Quality degradation during the content production phase Artefact introduced by lossy compression Network transmission errors Application and display device-specific parameters End-user’s preferences and perception model Kamil Kawa, Mikolaj Leszczuk, Atanas Boev Video Quality Assessment in Recognition Tasks 6 / 23
  • 10. www.agh.edu.pl Problem Introduction Problem Introduction Moreover, methods and metrics, to fulfil following expectations: In-service applicable Non-reference quality assessment High performance for diverse video content Coverage of all mentioned factors contributing to overall QoE Mapping between measured parameters (QoS, artefacts level) and QoE Kamil Kawa, Mikolaj Leszczuk, Atanas Boev Video Quality Assessment in Recognition Tasks 7 / 23
  • 11. www.agh.edu.pl Problem Introduction Problem Introduction Many parameters impacting ability to achieve recognition task, but selecting five of them as most important ones: 1 Usage time frame – specifying if one in need to analyse video in real-time or to be stored and analysed later 2 Discrimination level – specifying fine level of detail sought from video 3 Target size – specifying whether predicted region of interest in video occupies relatively small or large percentage of video 4 Lightning level – specifying anticipated lighting level of scene 5 Level of motion – specifying anticipated motion level in video scene Kamil Kawa, Mikolaj Leszczuk, Atanas Boev Video Quality Assessment in Recognition Tasks 8 / 23
  • 12. www.agh.edu.pl Assessments Environment Assessments Environment General viewing condition for subjective assessments to be met. Conditions divided into: Home environment Laboratory environment Kamil Kawa, Mikolaj Leszczuk, Atanas Boev Video Quality Assessment in Recognition Tasks 9 / 23
  • 13. www.agh.edu.pl Assessments Environment General Viewing Conditions for Subjective Assessments in Laboratory Environment General Viewing Conditions for Assessments in Laboratory Environment The assessors’ viewing conditions should be arranged as follows: Table: Viewing condition for subjective assessments in laboratory environment. Ratio of luminance of inactive screen to peak luminance: <=0.02 Ratio of the luminance of the screen, when displaying only black level in a completely dark room, to that corresponding to peak white: ≈ 0.01 Maximum observation angle relative to the normal 30 Ratio of luminance of background behind picture monitor to peak luminance of picture ≈ 0.15 Chromaticity of background D65 Other room illumination low Kamil Kawa, Mikolaj Leszczuk, Atanas Boev Video Quality Assessment in Recognition Tasks 10 / 23
  • 14. www.agh.edu.pl Assessments Environment General Viewing Conditions for Subjective Assessments in Home Environment General Viewing Conditions for Assessments in Home Environment Viewing distance and screen size to be selected in order to satisfy PVD (Preferred Viewing Distance). Table: Viewing condition for subjective assessments in home environment. Inactive screen vs. peak luminance <=0.02 Maximum relative vs. normal observation angle 30◦ Screen size for a 4/3 format ratio Screen size to satisfy PVD rules Screen size for a 16/9 format ratio Screen size to satisfy PVD rules Monitor processing without digital processing Peak luminance 200 cd/m2 Environmental illuminance on the screen 200lux Kamil Kawa, Mikolaj Leszczuk, Atanas Boev Video Quality Assessment in Recognition Tasks 11 / 23
  • 15. www.agh.edu.pl Assessments Environment General Viewing Conditions for Subjective Assessments in Home Environment General Viewing Conditions for Assessments in Home Environment PVD in function of screen size shown in Table Information in table and function recommending PVD related screen size that should be used Table: Information on PVD and related screen sizes. Screen diagonal (in) Screen height (H) PVD 4/3 ratio 16/9 ratio (m) (H) 12 15 0.18 9 15 18 0.23 8 20 24 0.30 7 29 36 0.45 6 60 73 0.91 5 >100 >120 >1.53 3-5 Kamil Kawa, Mikolaj Leszczuk, Atanas Boev Video Quality Assessment in Recognition Tasks 12 / 23
  • 16. www.agh.edu.pl Assessments Environment General Viewing Conditions for Subjective Assessments in Home Environment General Viewing Conditions for Assessments in Home Environment Figure: PVD for moving images. Kamil Kawa, Mikolaj Leszczuk, Atanas Boev Video Quality Assessment in Recognition Tasks 13 / 23
  • 17. www.agh.edu.pl Methods Methods ITU-T P.912 Recommendation introducing lot of useful methods for recognition tasks This Recommendation defining subjective assessment methods for evaluating quality of one-way video used for target recognition tasks Kamil Kawa, Mikolaj Leszczuk, Atanas Boev Video Quality Assessment in Recognition Tasks 14 / 23
  • 18. www.agh.edu.pl Methods Single Choice Method Single Choice Method Method to be used when single, unambiguous answer to identification question Technique utilisable for alphanumeric character recognition scenarios Experimenter asking tester which letter(s), or number(s) appearing in specific area of video Answer to be evaluated only as binary one: Correct Incorrect According to that, Yes or No test – also acceptable Kamil Kawa, Mikolaj Leszczuk, Atanas Boev Video Quality Assessment in Recognition Tasks 15 / 23
  • 19. www.agh.edu.pl Methods Single Choice Method Single Choice Method For example, one asking viewer if certain object present in scene In such a method – essential to ensure availability of easy to understand answers Care also to be taken to avoid terminology differences Use of “unsure” answer allowed, but not recommended Reason for that: over-usage of this option by testers, leading to contamination of experiment result Kamil Kawa, Mikolaj Leszczuk, Atanas Boev Video Quality Assessment in Recognition Tasks 16 / 23
  • 20. www.agh.edu.pl Methods Multiple Choice Method Multiple Choice Method This method especially appropriate for all discrimination class levels (introduced in ITU-T P.912) and target categories For this method, experimenter showing: Video (this slide) List of possible answers (next slide) After presenting video, viewers to choose label being closest to what recognised on clip Kamil Kawa, Mikolaj Leszczuk, Atanas Boev Video Quality Assessment in Recognition Tasks 17 / 23
  • 21. www.agh.edu.pl Methods Multiple Choice Method Multiple Choice Method Use of fixed multiple choices eliminating any possible misunderstanding possibly arising from open questions Due to that, more accurate measurements possible Number of choices offered to tester depending on number of alternative scenes presented As in previous method, one to take special care when ”unsure” is one of listed choices Kamil Kawa, Mikolaj Leszczuk, Atanas Boev Video Quality Assessment in Recognition Tasks 18 / 23
  • 22. www.agh.edu.pl Methods Timed Task Method Timed Task Method Viewer to be asked to watch for particular action or object that viewer about to recognise in video clip When tester perceiving target occurrence, pushing button In timed task, experimenter to determine whether time falling within acceptable time-frame for decision making This time-frames applicable for example in video scenarios: “A tester is responding to a violent situation and needs to identify whether crowd members have real weapons.” “A person is chasing a vehicle and needs to read the license plate.” Kamil Kawa, Mikolaj Leszczuk, Atanas Boev Video Quality Assessment in Recognition Tasks 19 / 23
  • 23. www.agh.edu.pl Methods Scenes Scenes TRV used to perform a recognition task Scenes to contain targets consistent with the application under study Measurement of test mostly focused on subject’s ability to identify objects and actions Problem: viewer possibly memorising scene content and using other visual clues to remember identity of target Therefore, set of scenes containing multiple versions to replace particular scene Kamil Kawa, Mikolaj Leszczuk, Atanas Boev Video Quality Assessment in Recognition Tasks 20 / 23
  • 24. www.agh.edu.pl Methods Scenes Scenes It is best way to control differences between versions Example scenario: “A person walks across the field of view carrying objects.” Set of videos to consist of multiple shots using different object and different people Number of scenarios in set to be large enough to reduce likelihood of scene memorisation Of course, experts to determine content of each scene Kamil Kawa, Mikolaj Leszczuk, Atanas Boev Video Quality Assessment in Recognition Tasks 21 / 23
  • 25. www.agh.edu.pl Methods Scenes Scenes Difference to be, for example, object carried by person on video Experts to identify critical tasks or parameters of scenes One to base set of multiple-choice answer and experiment design on these parameters One to create scene in way that, object of interest appearing in video at resolution realistically expected in practice Kamil Kawa, Mikolaj Leszczuk, Atanas Boev Video Quality Assessment in Recognition Tasks 22 / 23
  • 26. www.agh.edu.pl Conclusion Conclusion More general review of possible methods that could be used to video quality assessment presented in paper Based on limitation of QoS for video applications, QoE describing performance of whole end-to-end video delivery system from user’s point of view Key point for every test session: general viewing conditions for subjective assessments in laboratory and home environment Various methods applicable to recognition task defined in paper Each method containing list with brief description of used methods in scientific experiments Kamil Kawa, Mikolaj Leszczuk, Atanas Boev Video Quality Assessment in Recognition Tasks 23 / 23