Short answer is that this does not happen naturally in a relevant amount.
Photons of shorter wavelength carry more energy. So a one-photon process will always have an emission spectra at the same or lower frequency than the excitation frequency (see common fluorophores here). As the opposite would violate energy conservation (with no external energy source, like you asked).
Nonetheless, there are two-photon processes that can yield shorter wavelength photons, but these tend to be very small effects, lets look at rubidium for example:
You can see that it is possible to excite the atom to the 5D level with two IR photons to create a nice 420 nm photon. The problem is that the decay time of the intermediate level is very short (26 ns), so this will almost never happen unless you shoot lasers at it.