How to get `nvinfer` to be as accurate as TensorRT's API?

Accelerated Computing Intelligent Video Analytics DeepStream SDK
, , , , ,
7
  1. nvstreammux only accept NV12/RBGA/I420. In Deepstream test, there is no capsfilter after nvvideoconvert. nvvideoconvert maybe output NV12 format. please refer to this FAQ for how to dump pipeline graph. then you can check if there is data conversion, which will cause data loss.
  2. To reduce date conversion, please refer to thefollowing pipeline.
appsrc name=appsrc caps=video/x-raw,format=RGB ,width=128,height=128,framerate=0/1  ! nvvideoconvert !  video/x-raw\(memory:NVMM\),format=RGBA ! mux.sink_0 nvstreammux name=mux batch-size=1 width=128 height=128  ! nvinfer name=nvinfer config-file-path=config_nvinfer.yml  ! fakesink.