Proposal for a project to develop full reference media-layer objective metric for stereoscopic three-dimension video NTT Service Integration Laboratories,

Proposal for a project to develop full reference media-layer objective metric for stereoscopic three-dimension video

NTT Service Integration Laboratories, NTT CorporationTaichi Kawano

Jun Okamoto13 Jun., 2012

Background 2

Broadcasting

Many 3D services have been launched worldwide (as indicated in Table 1 in our contribution).

IPTV services

Several 3D services have started. NTT group is also providing 3D services on IPTV service.

3D video services become widespread and their markets are expected to expand worldwide.

Demand for 3D objective metric 3

Quality planning and monitoring are important tasks for service providers to provide high-quality service.

In IPTV service of NTT group, operators at head end check 2D video quality.

NTT recently received request from service provider of NTT group to automatically monitor 3D video quality because it is hard for operators to continually watch 3D videos.

Objective metric for 3D video quality “at head end” is necessary.

Proposal (1/3) 4

VQEG has been studying subjective assessment methods in order to update ITU-R Rec. BT.1438.

VQEG has not investigated objective metric for 3D services.

Objective metric needs to be developed for 3D services.

NTT proposes to launch new project to develop objective metric that will run in parallel with project for subjective assessment method.

Proposal (2/3) 5

Our focus is on developing objective metric for monitoring “3D video quality” “at head end” as “Phase I”. We divide objective metric for 3D services

into four groups by scale and monitoring point.

Monitoring point

Head end (Format and coding)

End user (Network)

Scale3D video quality Phase I Phase II

Other QoE factors(Depth, comfort, etc.) Phase II Phase II

Proposal (3/3) 6

FR media-layer objective metric is suitable in Phase I because highly accurate and codec-independent model is desired in head end.

Target applications 7

Target applications of objective metric for Phase I are as follows:

Quality monitoring for 3D video service at head end.

Quality measurement for 3D video encoding (e.g., for VoD and package media).

Why focus 3D video quality 8

Appropriate to focus on developing objective metric for “3D video quality” as Phase I

Many subjective assessment tests for 3D video quality have been already conducted.

Further study is needed to establish subjective assessment method for other QoE factors.

Other QoE factors will be (should be) different dimension from 3D video quality.

Reasonable to investigate objective metric for 3D video quality earlier than that for other QoE factors.

Processing chains for 3D services 9

(STEP A) Video format + encoding (STEP B) Delivering(STEP C)

Decoding + display

IP/Broadcastingnetwork

Server/Broadcasting tower

H.264/AVCencoder

3D videoSide by side (SbS)

TV

3D videoLeft/Right view

3D videoRight/Left view

H.264/MVCencoder+authoring

tool

Blu-rayplayer

1) SbS on conventional 2D delivery system

2) H.264/MVC on Blu-ray

TV

H.264/AVCencoder

3D videoLeft/Right view

STB

HGWIP networkServerH.264/AVC

encoder3D video

Right/Left view

TV antenna

Broadcastingnetwork

Broadcastingtower

3) 3D Simulcast

HGW/TV antenna

Blu-ray disc

Head end Network End user

TV(STB)

Why focus head end monitoring 10

Appropriate to focus on developing objective metric for “head end” as Phase I.

Adequate method is differ depending on monitoring point because technical requirement (e.g., target degradation, available information, and computational resource) is differ.

There is demand for automatic quality monitoring at head end.

Packetization schemes for some delivery systems have not been defined.

Conclusion 11

NTT proposed to launch new project to develop objective metric for 3D services.

Our focus is on developing “FR media-layer objective metric” for monitoring “3D video quality” “at head end”.

Supposed prodedure 13

STEP 1:Propose ToR and test plan (next meeting)

STEP 2:Propose CfP (first half of 2013)

STEP 3:Close CfP and conduct verification experiment (second half of 2013)

STEP 4:Analyze the results and prepare a report (2014)

Efficient to use part of test plan and database of subjective assessment test in 3DTV project.Supposed procedure is as follows.