fix receptive field comment (#2070)

This commit is contained in:
Adrià Arrufat 2020-04-27 19:02:26 +09:00 committed by GitHub
parent 8055b8d19a
commit 64ba66e1c7
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
1 changed files with 1 additions and 1 deletions

View File

@ -33,7 +33,7 @@ using namespace dlib;
// convolutionally over an entire image pyramid. Think of it like a normal // convolutionally over an entire image pyramid. Think of it like a normal
// sliding window classifier. This means you need to define a CNN that can look // sliding window classifier. This means you need to define a CNN that can look
// at some part of an image and decide if it is an object of interest. In this // at some part of an image and decide if it is an object of interest. In this
// example I've defined a CNN with a receptive field of a little over 50x50 // example I've defined a CNN with a receptive field of approximately 50x50
// pixels. This is reasonable for face detection since you can clearly tell if // pixels. This is reasonable for face detection since you can clearly tell if
// a 50x50 image contains a face. Other applications may benefit from CNNs with // a 50x50 image contains a face. Other applications may benefit from CNNs with
// different architectures. // different architectures.