Using Jetson Nano for custom object detection in real-time

Hello everyone,
I don’t know if its right platform for asking this but, I am an intern and I have created the Inference graphs for the custom object detection and I want to implement the same in Jetson nano. I have tried it with mobilenet ssd model on raspberry pi but it was very slow so I purchased jetson nano now. I am not sure if to follow the same steps as in raspberry or there is some simpler process of implementing the created inference graphs or to train on nano using its GPUs.

Please suggest as I don’t have much time for its implementation now.

thanks

Hi sauravkr28,

You could refer to below topic: https://devtalk.nvidia.com/default/topic/1055129/jetson-nano/imagenet-camera-cpp-libraries-not-working-on-jetson-nano/

Hope this can help.

Thanks

I @kayccc @sauravkr28 also have the same issue that I have yolov3 weights as well as tensorflow 2.0 weights and tflite weights with me but how can I use these weights with nano for infrence?

Hi farjadjaider3253,

Please open a new topic for your issue. Thanks

Ok @kayccc