5
$\begingroup$

Is there any literature about the different ways translation invariance can be achieved when classifying images with Convolutional Neural Networks? Aside from using the structure of CNN, did anyone attempt something different? Pre-processing, for example?

$\endgroup$
4
$\begingroup$

This answer by Matt Krause on What is translation invariance in computer vision and convolutional netral network? contain some pointers:

One can show that the convolution operator commutes with respect to translation. If you convolve $f$ with $g$, it doesn't matter if you translate the convolved output $f*g$, or if you translate $f$ or $g$ first, then convolve them. Wikipedia has a bit more.

One approach to translation-invariant object recognition is to take a "template" of the object and convolve it with every possible location of the object in the image. If you get a large response at a location, it suggests that an object resembling the template is located at that location. This approach is often called template-matching.

You may also find this technical report interesting, they give some overview: Leibo, Joel Z., Jim Mutch, Lorenzo Rosasco, Shimon Ullman, and Tomaso Poggio. "Learning generic invariances in object recognition: translation and scale." (2010). https://scholar.google.com/scholar?cluster=17887886525836197513&hl=en&as_sdt=0,22 ; http://cbcl.mit.edu/cbcl/publications/ps/Efficiency_of_invariance_and_learning_CBCL_TR.pdf

| cite | improve this answer | |
$\endgroup$
0
$\begingroup$

I can't give you a specific link, but I'd start looking into convolutional neural networks (CNNs). I don't know whether there are other approaches to this problem.

| cite | improve this answer | |
$\endgroup$
  • 2
    $\begingroup$ This isn't an answer to the question: OP has identified that CNNs are explicitly what s/he is asking after. $\endgroup$ – Sycorax says Reinstate Monica Sep 2 '16 at 14:25
  • 3
    $\begingroup$ He hadn't done so when I posted my answer $\endgroup$ – RdeWolf Sep 2 '16 at 15:05
  • $\begingroup$ That's correct. It's risky to answer a question that is vague or ambiguous, because you might misinterpret it (which obviously wouldn't be your fault). That's the reason why we have a comment mechanism for asking posters to clarify their questions. When you earn a little more reputation you will gain the ability to post such comments. (I have upvoted your answer here to help you on your way.) In the meantime it might work better to focus on well-posed questions or to wait until the community has helped improve the vaguer questions before you answer. $\endgroup$ – whuber Sep 2 '16 at 15:45

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.