Cascade more inference models in same file

I want to run models as cascade network like object detect model (detect person) >> face detection >> emotion classification in same file.
Has anyone can suggest me how to implement it? (I want to save result images in all step)



It’s recommended to check our Deepstream sample first.
Deepstream provide component based inference engine(TensorRT backend), which allows you to design a customized pipeline.


Thanks for your guide. After try to search,I got this repo that is good example.