0
$\begingroup$

In this lecture I don't understand how the output has 1 layer after applying the feature map to the 3 layer input.

$\endgroup$
4
  • 1
    $\begingroup$ is the feature map the 5*5*3 filter? If so applying a 3 layer filter to a 3 layer input would produce a 1 layer output wouldn't it? $\endgroup$ Jun 29, 2019 at 3:15
  • $\begingroup$ Yes, it is. my question is why it would turn to be 1 from 3 $\endgroup$ Jun 29, 2019 at 3:16
  • 1
    $\begingroup$ she says, "center the filter on each pixel and perform a dot product which gives one value per position. So I'm guessing it is a dot product between (L1, L2, L3) dot (F1, F2, F3) where L's are pixel values in the input and F's are whatever values are in the filter I think. $\endgroup$ Jun 29, 2019 at 3:23
  • $\begingroup$ Yes that makes sense now. Thanks for responding. $\endgroup$ Jun 29, 2019 at 3:25

1 Answer 1

1
$\begingroup$

In the video the speaker says, "center the filter on each pixel and perform a dot product which gives one value per position." So I'm guessing it is a dot product between (L1, L2, L3) dot (F1, F2, F3) where L's are pixel values in the input and F's are whatever values are in the filter.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.