Any machine learning engineer here please let me know how to deploy a model using mtcnn to extract features of a face then using vgg to identify faces . Anyone did this kind of deployment please let me know
b
broad-camera-13514
02/02/2021, 10:33 AM
Hey I'm an MLE. What framework are you using? The easies way to deploy is as a flask app.
o
orange-action-12733
02/02/2021, 11:39 AM
Hii sudharshan data science team developed a model which collect frames from image using mtcnn and then pass on to vgg for face detection they want this to be deployed. I am aware of single models this involves multi model so wondering how we do it
b
broad-camera-13514
02/02/2021, 1:24 PM
The same way. You can create a pipeline and deploy it. Flask is just a way to create a server out of a model or a pipeline so people can access it from outside.
The inputs to the app will the input frames and then feed to pipeline --> return the pipeline output as app output
o
orange-action-12733
02/02/2021, 2:07 PM
Okay since am newbee things seem complex I will reach you on zoom or skype one day and get in detail . Thanks for time and suggestions