Question: which techniques are preferred or popular in place recognition tasks?

I'm interested in building an autonomous mobile robot and I want it to recognize visited places. At this point I have found the vocabulary search tree and FAB-MAP but I don't know if those are widely used or are state of the art techniques