Does Speech enhancement of publicly available data help build robust Speech Recognition Systems?

Re-assign community