1.Thanks for your insights regarding the question I’ve stated before about multiple faces in the images. So what if someone takes a random input image of an actors(actors your model is been trained) and if the input image has two actors together, so that first it will find the faces using the code from your previous response , then it is fed into the model for classification ?.

So the output say it has 70% salman and 30% sharuk (if sharuk and salman exists in a single image which is fed as input) or it will give the probability of the highest just the “salman(70%) since we use softmax at the output layer.

2. And I’m just getting started into convolutional neural networks, the doubt regarding is that of image pixel. If we have a color image of 32x32 (total 1024 pixel values) and color images are represented using RGB values , so for each single pixel we have 3 inputs in a matrix [R_val , G_val , B_val].

If it is black and white , each pixel have values ranges from (0,1), so the output from each pixel is a 1D array [B_val or W_val].

Then how for a color image a kernel filter is first applied, what value it takes from the first pixel out of 3 values or it filters 3 times the whole image.

I know it sounds insane, but I was intimidated by it without knowing how it works. Can you shed some light ?

I don’t even know whether I am asking the correct question.

Thanks in advance.

Written by

Machine and Deep Learning Engineer

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store