1
$\begingroup$

I'm newbie in artificial intelligence.

I have started to research about how to do image segmentation and all the papers that I have found are about CNN. Most of them use the same network, U-NET, but with little variations: with more or less layers, different parameters values, etc.; but with not very different results.

It seems that CNN are in fashion and everyone use them. Or there are another reasons that I don't know.

If everyone is getting not very different results, why are they using the same approach instead of trying different ones?

$\endgroup$
1
$\begingroup$

CNN is used since, it is effectively an optimized use case for dealing with image data.

CNN effectively automatically extracts features from an images. Other techniques are more likely to not take full advantage of the data. CNN is able to make full use of the data by also including information from adjacent pixels and downsample through layers.

Paper on Performance of CNN on Image Data

Paper Comparing CNN's to other Methods

$\endgroup$
  • $\begingroup$ Maybe you should comment on the alternatives to CNNs and why CNNs are supposedly better than those alternatives. Maybe you should also specifically cite papers that support your statements. $\endgroup$ – nbro Jan 16 at 17:30
  • $\begingroup$ The reason I cited that particular paper was that it was more of an introductory paper. Given that the original poster seemed not too familiar, I thought it was appropriate. $\endgroup$ – Ryan Marinelli Jan 16 at 17:41
  • $\begingroup$ Thanks for your answer. If they are so good, why don't they reach 100% success? $\endgroup$ – VansFannel Jan 16 at 18:11

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.