Which loss function to implement for CNN-SVM infusion

Hi

I am using Matlab R2018b and am tryinbg to infuse svm classifier within CNN. My plan is to use CNN only as a feature extractor and use SVM as the classifier. Doing this, I got struck in a point during backward propagation. In this phase, I got puzzled as which loss function I need to implement to upgrade the gradients and the parametrers. Few points came up during this:

1. I got a feeling to implement the hinge loss here. But which form of hinge loss I will implement?(Should I move on to the second form of hinge loss imeplementation?)
2. As in Matlab R2018b, they have updated the parameters after calculating the gradients and forward loss(in their trainer file). If I would like to implement the second form of hinge loss, should i change the usual parameter update code of the Matlab R2018b?

Any form of advice doing this CNN-svm infusion will be appreciated as I am not finding any such material implemented in Matlab to get help.

thanks,