Face Recognition score

[maxwellnewage]

I use the following command: br -algorithm FaceRecognition -compare face1.png face2.png

But the result it's very confuse, i need a result like 0 to 1, but the output is -1 to 23.926773071289062.

[ghost]

Hi

On 28-Jul-2017 12:19 AM, "Maximiliano Burgos" notifications@github.com wrote:

I use the following command: br -algorithm FaceRecognition -compare face1.png face2.png

But the result it's very confuse, i need a result like 0 to 1, but the output is -1 to 23.926773071289062.

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/biometrics/openbr/issues/538, or mute the thread https://github.com/notifications/unsubscribe-auth/AOUh-9_THqhpyfoKTaB2pi0CcKUMOm_eks5sSNubgaJpZM4OlsJ1 .

[ghost]

Hi,

How eigenfaces are related to open BR. They were mentioned in the paper that they were using but I didn't come across this part in the documentation.

On 28-Jul-2017 12:19 AM, "Maximiliano Burgos" notifications@github.com wrote:

I use the following command: br -algorithm FaceRecognition -compare face1.png face2.png

But the result it's very confuse, i need a result like 0 to 1, but the output is -1 to 23.926773071289062.

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/biometrics/openbr/issues/538, or mute the thread https://github.com/notifications/unsubscribe-auth/AOUh-9_THqhpyfoKTaB2pi0CcKUMOm_eks5sSNubgaJpZM4OlsJ1 .

[maxwellnewage]

sorry but i don't understand, can you explain me more?

[bklare]

Eigenfaces operates on the pixels of loosely aligned face images. OpenBR's algorithms first "tightly" aligns the individual facial components. Next, they are represented with local descriptors instead of pixels. Finally, LDA is used instead of (or, more precisely, in conjunction with) PCA.

[maxwellnewage]

mm ok, so how can validate the similarity faces? Because i understand the theory, but the result number of command line:

br -algorithm FaceRecognition -compare 2017-07-20-184437.jpg 2017-07-20-184437.jpg
Set algorithm to FaceRecognition
Loading /usr/local/share/openbr/models/algorithms/FaceRecognition
Comparing 2017-07-20-184437.jpg and 2017-07-20-184437.jpg
Enrolling 2017-07-20-184437.jpg to 2017-07-20-184437m74zbc.mem
100.00%  ELAPSED=00:00:00  REMAINING=00:00:00  COUNT=1
100.00%  ELAPSED=00:00:00  REMAINING=00:00:00  COUNT=1
23.9268

I don't understand the meaning of the number 23...

[ghost]

Refer open br research paper

On 28-Jul-2017 9:12 PM, "Maximiliano Burgos" notifications@github.com wrote:

mm ok, so how can validate the similarity faces? Because i understand the theory, but the result number of command line:

br -algorithm FaceRecognition -compare 2017-07-20-184437.jpg 2017-07-20-184437.jpgSet algorithm to FaceRecognition Loading /usr/local/share/openbr/models/algorithms/FaceRecognition Comparing 2017-07-20-184437.jpg and 2017-07-20-184437.jpg Enrolling 2017-07-20-184437.jpg to 2017-07-20-184437m74zbc.mem 100.00% ELAPSED=00:00:00 REMAINING=00:00:00 COUNT=1 100.00% ELAPSED=00:00:00 REMAINING=00:00:00 COUNT=1 23.9268

I don't understand the meaning of the number 23...

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/biometrics/openbr/issues/538#issuecomment-318688135, or mute the thread https://github.com/notifications/unsubscribe-auth/AOUh-9vw6m0fvrG-bWDHHDyWnaR9cWB8ks5sSgF9gaJpZM4OlsJ1 .

[maxwellnewage]

where can i find that paper?

[ghost]

Website itself you will find. I want to know how face recognition work other than just comparing two images using already existing model.

On 28-Jul-2017 11:52 PM, "Maximiliano Burgos" notifications@github.com wrote:

where can i find that paper?

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/biometrics/openbr/issues/538#issuecomment-318727641, or mute the thread https://github.com/notifications/unsubscribe-auth/AOUh-4nuB1VyTlTywVmRoAIAlseh9I_6ks5sSibhgaJpZM4OlsJ1 .

[maxwellnewage]

following this post => https://groups.google.com/forum/#!topic/openbr-dev/oXYXC2WvMaw The result was 0-1 but in the code the output was 6000 to 7000, strange. This is my code:

float FaceRecognition::addImgs(QTemporaryFile* imgFrom, QTemporaryFile* imgTarget)
{
    QSharedPointer<br::Transform> transform = br::Transform::fromAlgorithm("FaceRecognition");
    QSharedPointer<br::Distance> distance = br::Distance::fromAlgorithm("FaceRecognition");

    if(!imgFrom->open())
        return 0;

    if(!imgTarget->open())
        return 0;

    QString filePathFrom = imgFrom->fileName();
    QString filePathTarget = imgTarget->fileName();

    br::Template queryA(filePathFrom);
    br::Template target(filePathTarget);

    queryA >> *transform;
    target >> *transform;

    float comparison = distance->compare(queryA,target);

    return comparison;
}