An objective technique for evaluating doubletalk detectors in acoustic echo cancelers

Echo cancelers commonly employ a doubletalk detector (DTD), which is essent ial to keep the adaptive filter from diverging in the presence of near-end speech and other disruptive noise. There have been numerous algorithms to d etect doubletalk in an acoustic echo canceler (AEC), In those applications, typically, the threshold is chosen only by some heuristic method and the p erformance evaluation is very subjective. In this study, we develop a way t o objectively evaluate DTD algorithms based on the standard statistical met hods of detection theory. A receiver operating characteristic (ROC) is deri ved to characterize DTD performance. Several DTD algorithms are examined an d simulated under typical real-world operating conditions using measured ro om responses and signals taken from a digital speech database. The DTD meth ods are then evaluated and compared using the ROC metric.