Ingrid Jafari, Serajul Haque, Roberto Togneri, Sven Nordholm: On the integration of time-frequency masking speech separation and recognition in underdetermined environments. ACSCC 2012: 1613-1617