Problem to create engine from SSD UFF on Jetson Nano (Jetbot)

I have created my model using tensorflow object detection API, so I have my pb models.

then i create the UFF model by using this function:

ff_model = uff.from_tensorflow_frozen_model("./frozen_inference_graph.pb", output_name)

Then I want to create the Tensor RT engine using this function:

engine = ssd_tensorrt.ssd_uff_to_engine(uff_buffer)

but I get error lambda missing argument network. I have checked the code of ssd_uff_to_engine, it is written that the network has been defined by the code.

So what is wrong?

Hi,

Please refer below example for converting SSD-Mobilenet from Tensorflow to TensorRT:

https://github.com/AastaNV/TRT_object_detection
https://jkjung-avt.github.io/tensorrt-ssd/
https://jkjung-avt.github.io/hand-detection-tutorial/

Thanks