Op type not registered 'TFLite_Detection_PostProcess' for face ssd mobile v2

My system config:
cuda 10.0
cudnn 7.6
ubuntu 18.04
Tensorflow 1.15.0 binary
GPU : GTX 1080

I want to run the facessd_mobilenet_v2_quantized_open_image_v4, I can't load the .pb model graph and I get this error:

tensorflow.python.framework.errors_impl.NotFoundError: Op type not registered 'TFLite_Detection_PostProcess' in binary running on mm-pc. Make sure the Op and Kernel are registered in the binary running in this process. Note that if you are loading a saved graph which used ops from tf.contrib, accessing (e.g.) tf.contrib.resampler should be done before importing the graph, as contrib ops are lazily registered when the module is first accessed.

I used this simple code for load graph:

    model = tf.Graph()
    with model.as_default():
        model_graph_def = tf.GraphDef()
            with tf.gfile.GFile(pb_path, "rb") as f:
                serialized_graph = f.read()
                model_graph_def.ParseFromString(serialized_graph)
                tf.import_graph_def(model_graph_def, name="")
            config = tf.ConfigProto()
            config.gpu_options.allow_growth = True
            model_sess = tf.Session(graph=model, config=config)

I get error in this line

tf.import_graph_def(model_graph_def, name="")