How do I convert a Net to VINO when both .pb and pbtxt format are used to read the net - which of the two best serves ?
frozen_graph = str("detection/240x180_depth0.75_ssd_mobilenetv1/frozen_inference_graph.pb")
text_graph = str("detection/240x180_depth0.75_ssd_mobilenetv1/graph.pbtxt")
cvNet = cv2.dnn.readNetFromTensorflow(frozen_graph, text_graph)
Which of the .pb and pbtxt do I use above as their nodes contents are different?
Dearest K, Mike:
You can convert a text or binary version of a Tensorflow protobuf (or in the case of binary, meta and checkpoint files) to a frozen protobuf using Tensorflow tools. From Model Optimizer's perspective, it doesn't care as long as the Tensorflow Protobuf you feed in is frozen. In fact mo_tf.py has a switch --input_model_is_text.
Please kindly review the following mo tf doc .