POLQA reference C++ code is rubbish

If you are researchers who would like to experience the POLQA algorithms specify in ITU-T P.863, your will be very upset!
The source codes given in [You must be registered and logged in to see this link.] can not be compile in anyway, and there are abundant of mistakes made in ! The coding style of the programmers are terrible, and even a high school student can do it better. IMO, the ITU and these companies
OPTICOM GmbH, Erlangen, Germany
SwissQual AG, Solothurn, Switzerland
TNO Telecom, Delft, The Netherlands
KPN, The Hague, The Netherlands
are irresponsible and shameless to publish the rubbish codes on their Website. They should be condemned.

Hi,

Why are you trying with POLQA? Please take a look at Sevana AQuA that works well instead of PESQ and POLQA for many companies already.

Thank you for noticing my post， indeed I have come across Sevana‘s products but not impressed by it. I'm not sure if it is a reliable company, since I can not catch the idea of the test results at a first glance. And a post of my friend on the web said it is a small company with about 10 emploees. [You must be registered and logged in to see this link.] it with Google if you cannot read chinese).

Thank you for your suggestion,I will have a try with their product.

Thanks for the review! It's quite interesting... I think the main problem is that Sevana web site is outdated, but then just check list of customers and try to contact some of them for independent reference. Ask Sevana for references and finally get evaluation and make up your mind yourself. Like they say in Scandinavia it's not a matter of quantity, but a matter of quality The smaller the company the better customer service you get

Hello, Could anyone send me a copy of AQuA project, it can't be download anymore!!
and did anyone figure out how to compile the POLQA algorithm?
Thanks!!

You can apply a time-limited free evaluation edition ，for more information please go to their official website.

Since the POLQA does not provide any evaluation tool nor test vectors even for academic researchers,  I think this standard should be labeled as "Unavailable"!  In fact,  objective audio/speech quality measurement is no more than a mapping from a distortion measurement to a subjective MOS.  By providing a widely acknowledged database including audio/speech testing vectors and their corresponding MOS values, just like ITU did for G.729,  it is quite easy to use modern statistical learning algorithms to pursuit this mapping, which means everyone can develop his/her own quality evaluation method.  In a general sense,  with sufficient training data all these methods will converged to the same solution. There is actually no need to encourage everybody to follow specific procedures.

halala wrote:If you are researchers who would like to experience the POLQA algorithms specify in ITU-T P.863, your will be very upset!
The source codes given in [You must be registered and logged in to see this link.]  can not be compile in anyway, and there are abundant of mistakes made in ! The coding style of the programmers are terrible, and even a high school student can do it better.  IMO, the ITU and  these companies
                OPTICOM GmbH, Erlangen, Germany
SwissQual AG, Solothurn, Switzerland  
TNO Telecom, Delft, The Netherlands                                
KPN, The Hague, The Netherlands      
are irresponsible and shameless to publish the rubbish codes on their Website. They should be condemned.