Running only classifier as primary, possible?

Please provide complete information as applicable to your setup.

• Hardware Platform (Jetson / GPU) - dGPU
• DeepStream Version - 5.1
• NVIDIA GPU Driver Version (valid for GPU only - GTX 1080)


I have made a python Deepstream app which takes in image input runs through primary LPDNet giving the number plate and then to secondary LPRNet giving OCR metadata, this was made by taking a reference sample app in cpp and blog post and implemented in python.

Now I have folder which contains cropped license plate images and just need to run LPRNet on it and get metadata, so can we only run classifier in Deepstream.

Is that possible ? if so than is there an python or cpp sample for it.


It’s possible. you can just refer deepstream-test1 and replace the model with your classifier model

thanks for the input
I was able to run deepstream-app3 as only classifier by copying the sgie config to pgie and making the changes in code i.e removing sgie and keeping only pgie, and in order to fetch meta took reference from test-app2.

The main changes to be made in config id process-mode=1 and network-type=1

This topic was automatically closed 60 days after the last reply. New replies are no longer allowed.