"Dinosaur", when used properly, refers to a clade. There's no way for a clade to evolve multiple times, or even to really evolve at all as something that happens; it's just a name for a group of descendants sharing a common ancestor.
I'm guessing instead that it's an incorrect transcription of "feathers in dinosaurs", since in and and sound quite similar out of context; that is, that something structured like a feather evolved three times in the dinosaur lineage.
I know little to nothing about this area, but Wikipedia's article on feathered dinosaurs suggests that this is a bit unresolved:
It is not known with certainty at what point in archosaur phylogeny the earliest simple "protofeathers" arose, or whether they arose once or independently multiple times.
There are feathers observed in multiple lineages of dinosaurs, but it seems most plausible that they all came from the same origin rather than independently. For the intervening species it may be that either they lost the feathered phenotype or that there is simply not fossil evidence of feathers for those species.
As for bipedalism, well, we don't really need any feathers to reason that a trait could evolve more than once. Flight itself is a simple example: there are flying insects, flying birds, and flying mammals (bats), all who evolved flight from previously flightless ancestors.
The most direct example, though, is bipedalism itself: there are many bipedal dinosaurs, including modern birds. Human bipedalism evolved separately from non-bipedal mammals. Arguing whether it's happened multiple times among the apes is another story, that has little to do with any examples from other clades or other traits except analogy. Certainly, other apes locomote on two legs at times, but none of them walk quite the way humans do. If you want to argue that feather evolution involved an earlier "protofeather" that then separately evolved into a full-fledged (hehe) version multiple times, then maybe that would relate to a "proto-bipedalism" (as observed in apes) that could result in multiple "truly bipedal" lineages, but one would need further evidence to support that.