
Hybrid audio call, April 29 2014
=======================================

Participants: Silvio, Chulhee, Christian, Werner, Alex, Margaret, Roland


Agenda:
-------------

Final report:

1. Does anyone need more time to check the ILG data analysis?  

2. Fill in data into report?

3. Post drafts on the Hybrid website?

4. Executive summary


Call:
--------------


1. Yonsei: Checked RMSE and Correlation starting from the coefficients: It is ok.
   Opticom: Some polynomials are not monotonically increasing (which ones?), but difference is not important.
            Otherwise numbers seem OK.
   SwissQual: OK for individual data sets (RMSE, corr), but did not check the superset.
   TLabs: Did some checking of individual data sets, but did not check the superset.

   Margaret: Will re-check the values for the super-set.

2. OK to put in data to the report.

3. Decision: Draft final report will be distributed by email.

4. Executive summary:
   
   Missing: Add scale of the scores, e.g. [1,5]. 
   Add impairments (encoders),... 

   Opticom: Would like to have more statistics in the executive summary, not only average values.
   
   Proposition: Put a short description of each data set into the executive summary.
   Proposition: Note some of the problems of the data sets, e.g. un-even distribution for some of the data sets.
   Proposition: Add per-data set correlation values to the executive summary.

   Margaret will update report including the above propositions before next weeks audio call.

   Silvio: re-read report and check for un-used text.

   Margaret will write email to each lab asking for description of their data set, including a short summary for the
   executive summary.



Next audio call: May 6, 16:00 CEST

