Question about 3 in "The Basics of Covnets" Quiz

Course Q&A Deep Learning Specialization Convolutional Neural Networks
1

Hello,
I missed 3 questions on the Basics of Convets quiz and I’m hoping someone can explain to me why I got them wrong.

For question 3, I calculated the number of parameters as 5 times 5 for the dimensions of the filter + 1 for the bias parameter for a total of 2600. The quiz told me that that each filter has a volume where “each filter is a volume where the number of channels matches up the number of channels of the input volume”, but in the lecture “Why Convnets” at timestamp 2:19 professor Ng computes the number of parameters in a layer with 6 5x5 filters and comes up with the total number of parameters as (5*5 +1) *6 = 156. He doesn’t take into account the number of channels in the input layer. Can someone please explain this discrepancy to me?

Update: I went back and watched the “Convolutions Over Volume” video and it says you do need to take into account the number of channels in the input layer. Is there a distinction between the situations being described in these two videos?

2

I think it must be the case that in the initial “Why Convnets?” video, Prof Ng is keeping it simple by talking about a single channel input. The later information which says you need to take the input channels into account is correct, of course.

To give a concrete example, if you have inputs that have 3 channels, your filters are 5 x 5, and you have 6 output channels, the number of parameters is:

(5 * 5 * 3 + 1) * 6 = 456