Intel® Distribution of OpenVINO™ Toolkit
Community assistance about the Intel® Distribution of OpenVINO™ toolkit, OpenCV, and all aspects of computer vision-related on Intel® platforms.
6574 Discussions

How to use less than 3 channel input signals like heat image of sound spectrum, gray scale

idata
Employee
956 Views

Hi,

 

I am using stick for the purpose of sound classification by using heat image of its frequency.

 

This image contains 2 channel ones which indicate stereo sound, but I don't understand how to change input number instead of 3 channel RGB.

 

Is it possible for me to change the input number down to less than 3?

 

Thanks.
0 Kudos
1 Reply
idata
Employee
657 Views

@KKK Yes, it should be possible to change the input number to less than 3. In your prototxt file, there should be an input header at the top that looks like:

 

input_shape {

 

dim: 1

 

dim: 3

 

dim: 227

 

dim: 227

 

}

 

The first dim represents the batch size (currently the NCS only supports a batch size of 1). The second dim is the number of channels and can be changed to suit whatever input the network requires.

0 Kudos
Reply