6
$\begingroup$

I have $m$ labeled images, each with 224x224 pixels and 5 different image channels. What is the best way to train a CNN architecture using this data when $m$ is small (less than 2000)? Is it possible to consider each channel as a separate image or is it better to input all $n$ channels into the input layer at the same time?

$\endgroup$
7
$\begingroup$

You should work with each image as a volume of dimensions 224x224x5. You still do 2D convolution over the first 2 dimensions as usual, but keep the entire 3rd dimension. For instance, if you use a 7x7 convolution window, each filter will produce a 224x224x1 volume as output (with stride = 1 and zero padding), and the convolutional layer as a whole will produce a 224x224xN volume, where N is the number of filters. ConvNetJS and many other conv nets libraries take this approach.

$\endgroup$
1
$\begingroup$

Regarding your second question:

What is the best way to train a CNN architecture using this data when mm is small (less than 2000)?

There are some common ways to artificially increase the training set size, e.g. adding jitter or doing some transformation such as rotations.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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