0
$\begingroup$

Meta-learning has 3 broad approaches: model, metric and optimization-based approach. Each of them has its own sub-approach, like matching network, meta-agonistic and Siamese-based network, and so on.

How do I decide which approach to select for a task? For my case, I have a noisy image, and they need to be compared with 10 different new images every time. Do I have to start with the trial and error method, or there is some methodology behind this approach selection?

Improve this question
$\endgroup$
1
  • $\begingroup$ Could you please provide a reference where you read about these "meta-learning approaches"? $\endgroup$
    – nbro
    Mar 9, 2021 at 10:03

1 Answer 1

Reset to default
0
$\begingroup$

The only principled approach is to see if the problem you are working has been worked on before. If yes, then, use what they used for your problem. It is the alternative to go through all the models one-by-one.

Like, in your case, it bears a semblance with face models, so, Siamese network with ArcFace loss for getting your embeddings is a good bet for you.

Improve this answer
$\endgroup$

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .