Weekly meeting 18 May 2016

Things I have done:

– Refactor the code of the cnn_test. Solve problems with using gift-grab dynamically linked, embedded in the code now).

– Still solving synchronisation bugs.

Ideas from Tom for creating a discriminative and data-driven feature descriptor:

– Create two networks that extract a feature vector from each image and use a combined loss so that it learns what good and bad matched are.

– The 2 network thing seems to be called a Siamese network, here is an example:
http://imagine.enpc.fr/~zagoruys/deepcompare.html