Let $X$ be a curve of genus at least two. Then one can associate an abelian variety to $X$; this is the Jacobian.
Let $X\to Y$ be a double cover of curves. Then we can associate an abelian variety to $X\to Y$; this is the Prym variety.
To a K3 surface one can associate an abelian variety; this is the Kuga-Satake variety.
Question. Can we associate a K3 surface to a curve (or just some special class of curves) in a reasonable/useful way?
Question 2. Can we do the same with $X\to Y$ covers of degree $d$ (or just degree two)?
This question is vague. I understand. I'm really just asking whether there is some way to go from a curve of genus at least two to a K3 surface.
Elliptic curves are abelian varieties. So for a genus one curve $E$ over $\mathbf C$ one could ask whether there is a K3 surface associated to it. I think in this case you could maybe write down an elliptic K3 surface $\mathcal E\to \mathbf P^1$ whose generic fibre is $E$. Still need to figure out if this is useful though...