Deploy tensorflow model on deep stream

Accelerated Computing Intelligent Video Analytics DeepStream SDK
1

I want to use .pb model file with deepstream. Is there a way to integrate .pb file or use it with deepstream?
I am using :
Jetson nano
deepstream5
Jetpack2.2
tensorRT7…0

3

Hi,

Deepstream uses TensorRT as the inference engine.
TensorRT currently supports Caffe, uff, and onnx format.

So please convert your .pb model into uff or onnx format first.
Here is a useful convert for your reference:

Thanks.