Revision history [back]

you need:

• a c++ compiler
• cmake
• the apache ant tool
• java jdk
• https://github.com/berak/opencv_contrib/blob/master/README.md

you need:

• a c++ compiler
• cmake
• the apache ant tool
• java jdk
• https://github.com/berak/opencv_contrib/blob/master/README.mdthe opencv_contrib repo, build instructions here

Dnn net = dnn.readNetFromTorch("openface.nn4.small2.v1.t7");
Mat inputBlob = dnn.blobFromImage(image, 1./255, Size(96,96), Scalar(), true, false);
net.setInput(inputBlob);
feature = net.forward();

you need:

• a c++ compiler
• cmake
• the apache ant tool
• java jdk
• the opencv_contrib repo, build instructions here

note, that there is also an easy to use pretrained openface cnn, cnn (in the Dnn module), you could use it like this:

Dnn net = dnn.readNetFromTorch("openface.nn4.small2.v1.t7");
Mat inputBlob = dnn.blobFromImage(image, 1./255, Size(96,96), Scalar(), true, false);
net.setInput(inputBlob);
feature = net.forward();

then you can compare features obtained above with a plain L2 norm.

(note, that this will also work oncolor images !)

you need:

• a c++ compiler
• cmake
• the apache ant tool
• java jdk
• the opencv_contrib repo, build instructions here

note, that there is also an easy to use pretrained openface cnn (in the Dnn module), you could use it like this:

Dnn net = dnn.readNetFromTorch("openface.nn4.small2.v1.t7");
Mat inputBlob = dnn.blobFromImage(image, 1./255, Size(96,96), Scalar(), true, false);
net.setInput(inputBlob);
feature = net.forward();

then you can compare features obtained above with a plain L2 norm.

(note, that this will also work oncolor images !)