Can I load in DeepLens two models at the same time

I am trying to load two MXNet well-trained models into one project. My goal is first detecting the objects and recognized them by the "object-detection" model. Then if there exits "person" in it and I will let the frame predicted by another "face-recognization" model to located the human faces. Otherwise it will pass the face detection model directly.

So my goal is to load two different models into a Deeplens Device. But when I operated on the project manage website
(DeepLens>Projects>Object-detection>Edit>Add Model),
I found out it can only load one model at the same time.
I am wondering if I can load two model at the same time ? Or I should load them manully put the model into the deeplens locally not by the website. Or it is totally not working.

Regards!
R

Edited by: RuiAcc on Nov 12, 2020 12:21 AM