Ask Your Question

Revision history [back]

Setting Input Parameters dnn.detection model

Hi all,

I have learnt opencv primarily through tutorials on Youtube.

One thing I often find frustrating is that many important steps such as setting parameters for your dnn: net.setInputSize(320,320) net.setInputScale(1.0/127.5) net.setInputMean((127,127,127)) net.setInputSwapRB(True)

are often overlook or not explained. I have tried iterating with different values to test performance, or reading through documentation but information in this area is sparse.

Where can I learn about the purpose and value from using such parameters.

If you search setting input mean dnn opencv, there is hardly any information.

The documentation also contains limited information on methods and classes too.

Thanks

click to hide/show revision 2
None

updated 2020-11-16 09:24:43 -0600

berak gravatar image

Setting Input Parameters dnn.detection model

Hi all,

I have learnt opencv primarily through tutorials on Youtube.

One thing I often find frustrating is that many important steps such as setting parameters for your dnn: dnn:

net.setInputSize(320,320)
net.setInputScale(1.0/127.5)
net.setInputMean((127,127,127))
net.setInputSwapRB(True)

net.setInputSwapRB(True)

are often overlook or not explained. I have tried iterating with different values to test performance, or reading through documentation but information in this area is sparse.

Where can I learn about the purpose and value from using such parameters.

If you search setting input mean dnn opencv, there is hardly any information.

The documentation also contains limited information on methods and classes too.

Thanks