the secondary infer results attached to the streaming showing on the screen in "deepstream-app"

I run the “deepstream-app” on V100 in a docker (nvcr.io/nvidia/deepstream 4.0.1-19.09-devel ), and use the configuration of the car detection you offered, which has one primary detector and there secondary classifier. I wonder that how does the secondary infer results attached to the streaming showing on the screen. I have scanned the code, but do not get it.
the pipeline shows below.

Hi
Check this,
/sources/gst-plugins/gst-nvinfer/gstnvinfer.cpp:1388 static GstFlowReturn
gst_nvinfer_process_objects (GstNvInfer * nvinfer, GstBuffer * inbuf,
NvBufSurface * in_surf)

you can also refer to documentation https://docs.nvidia.com/metropolis/deepstream/plugin-manual/index.html#page/DeepStream_Plugin_Manual%2Fdeepstream_plugin_details.02.01.html%23wwpID0E0IZ0HA

Thank you for your reply, I will work it out.