I'm not 100% sure on the exact mechanism, but I believe the basic idea is that tin is your lewis acid. It coordinates to the acetyl group's carbonyl oxygen, thus making it a better leaving group. The oxygen in the sugar ring then uses its lone pairs to kick out the acetyl group, thus making an oxocarbenium ion (which is just a better resonance description of a carbocation adjacent to an oxygen, BTW). The Me3SiN3 is just a source of azide, and the azide just does nucleophilic attack on the oxocarbenium ion.