I wouldn't exclude the possibility, but I imagine this to be pretty hard to develop. In the end, STED is based on emissions in the visible range and those usually provide a much more sensitive tool than Raman observations; in Raman, you have a lot of "side processes", the most dominant of which is of course the Rayleigh line. So, in order for this to be effective, I think you would have to overcome a huge sensitivity barrier.
Edit: I just found this, I think it comes close to what you have in mind:
Article Super-resolution vibrational microscopy by stimulated Raman ...