def inception_batch_inference():
tensors = freeze.unfreeze_into_current_graph(
paths.IMAGENET_GRAPH_DEF,
tensor_names=[consts.INCEPTION_INPUT_ARRAY_TENSOR, "softmax:0"])
def forward(sess, _img):
label_vect = sess.run(
tensors["softmax:0"],
{tensors[consts.INCEPTION_INPUT_ARRAY_TENSOR]: _img})
return label_vect
return forward
def inception_inference():
tensors = freeze.unfreeze_into_current_graph(
paths.IMAGENET_GRAPH_DEF,
tensor_names=[consts.INCEPTION_INPUT_STRING_TENSOR, "softmax:0"])
def forward(sess, image_raw):
label_vect = sess.run(
tensors["softmax:0"],
{tensors[consts.INCEPTION_INPUT_STRING_TENSOR]: image_raw})
return label_vect
return forward
def inception_model():
tensors = freeze.unfreeze_into_current_graph(
paths.IMAGENET_GRAPH_DEF,
tensor_names=[
consts.INCEPTION_INPUT_TENSOR, consts.INCEPTION_OUTPUT_TENSOR
])
def forward(sess, image_raw):
out = sess.run(tensors[consts.INCEPTION_OUTPUT_TENSOR],
{tensors[consts.INCEPTION_INPUT_TENSOR]: image_raw})
return out
return forward
def inception_image_model():
tensors = freeze.unfreeze_into_current_graph(
paths.IMAGENET_GRAPH_DEF,
tensor_names=[
consts.INCEPTION_INPUT_ARRAY_TENSOR, consts.INCEPTION_OUTPUT_TENSOR
])
def forward(_sess, _image_array):
_out = _sess.run(
tensors[consts.INCEPTION_OUTPUT_TENSOR],
{tensors[consts.INCEPTION_INPUT_ARRAY_TENSOR]: _image_array})
return _out
return forward