A Constrained pseudorandom function (CPRF) is a pseudorandom function that
provides fine-grained access to the evaluation of the function. In other
words, for a (possibly super-polynomial) subset of inputs, a constrained
pseudorandom function allows us to derive a constrained key that enables
evaluating the function on that subset of inputs while learning nothing
beyond. We propose and analyze a simple strategy for constructing 1-key
constrained pseudorandom functions from homomorphic secret sharing (HSS)
protocols. This relation, in particular, leads to instantiations of CPRFs
for various constraints and from various assumptions. In this talk, I
present this strategy and briefly go over one of the instantiations.