def __init__(self, model, iterator, alpha=0.2, verbose=2):
self.model = init(model)
self.iterator = iterator
self.alpha = alpha
self.verbose = verbose
y_tr = self.model[-1].op({'dropout':True, 'bn_active':True, 'infer':False})
y_te = self.model[-1].op({'dropout':False, 'bn_active':False, 'infer':False})
y_inf = self.model[-1].op({'dropout':False, 'bn_active':True, 'infer':True})
self.X = self.model[0].X
anchor = y_tr[::3]
pos = y_tr[1::3]
neg = y_tr[2::3]
dpos = pair_euclidean(anchor, pos)
dneg = pair_euclidean(anchor, neg)
d = dneg - dpos
cost = T.maximum((1. - self.alpha) - d, 0.).mean()
self.updates = collect_updates(self.model, cost)
self.infer_updates = collect_infer_updates(self.model)
self.reset_updates = collect_reset_updates(self.model)
self._train = theano.function([self.X], cost, updates=self.updates)
self._transform = theano.function([self.X], y_te)
self._infer = theano.function([self.X], y_inf, updates=self.infer_updates)
self._reset = theano.function([], updates=self.reset_updates)
def __init__(self, model, iterator, alpha=0.2, verbose=2):
self.model = init(model)
self.iterator = iterator
self.alpha = alpha
self.verbose = verbose
y_tr = self.model[-1].op({
'dropout': True,
'bn_active': True,
'infer': False
})
y_te = self.model[-1].op({
'dropout': False,
'bn_active': False,
'infer': False
})
y_inf = self.model[-1].op({
'dropout': False,
'bn_active': True,
'infer': True
})
self.X = self.model[0].X
anchor = y_tr[::3]
pos = y_tr[1::3]
neg = y_tr[2::3]
dpos = pair_euclidean(anchor, pos)
dneg = pair_euclidean(anchor, neg)
d = dneg - dpos
cost = T.maximum((1. - self.alpha) - d, 0.).mean()
self.updates = collect_updates(self.model, cost)
self.infer_updates = collect_infer_updates(self.model)
self.reset_updates = collect_reset_updates(self.model)
self._train = theano.function([self.X], cost, updates=self.updates)
self._transform = theano.function([self.X], y_te)
self._infer = theano.function([self.X],
y_inf,
updates=self.infer_updates)
self._reset = theano.function([], updates=self.reset_updates)
def PairEuclidean(y_true, y_pred):
return pair_euclidean(y_pred, y_true).mean()
def PairEuclidean(y_true, y_pred):
return pair_euclidean(y_pred, y_true).mean()
