TensorRT optimisation

I have trained a tensorflow 2 object detection model on my custom data set
Is it possible to convert tensorflow 2 object detection model to TensorRT for jetson nano

Hi,

Yes. The workflow looks like this: TensorFlow -> ONNX -> TensorRT.

1. You can convert a TensorFlow model into ONNX format with tf2onnx:

2. Then you can deploy the ONNX model with trtexec directly.

/usr/src/tensorrt/bin/trtexec --onnx=[your/model]

Thanks.

Hi,
Thank you for your time.
Is there any tutorial or a blog that can help me deploy tensorflow2 object detection models on jetson nano (with maximum efficiency)
An example of deploying tensorflow2 object detection model on jetso nano would be really helpful

Hi,

Please check this blog :

Thanks.