Open cheer37 opened 8 years ago
If you've already trained an embedding, you can write a test for it yourself using the Python interface.
Also, have you looked at the outstanding PRs for triplet nets (https://github.com/BVLC/caffe/search?q=triplet&type=Issues&utf8=%E2%9C%93)?
There is no triplet loss layer in caffe? How can i test my triplet loss embedding using caffe?