is dnn module possible to get specific outputs from multiple layers in one forwarding?

asked 2019-07-29 23:56:09 -0500

updated 2019-07-30 06:43:21 -0500

berak gravatar image

i.g.

 net = cv.dnn.readNet(caffemodel, prototxt)
 img = ...
 blob = cv.dnn.blobFromImage(img, scale, (227, 227), mean, isrgb, cropFromCenter)
 net.setInput(blob)

 A, B = net.forward([ "A", "B" ])
edit retag flag offensive close merge delete

Comments

LBerger gravatar imageLBerger ( 2019-07-30 03:01:27 -0500 )edit

yes, it would probably work like above, so where exactly is the problem ?

do you have a problematic model, you can show us ?

berak gravatar imageberak ( 2019-07-30 06:57:49 -0500 )edit