Model compensation techniques for noise-robust speech recognition approximate the corrupted speech distribution. This paper introduces a sampling method that, given speech and noise distributions and a mismatch function, in the limit calculates the corrupted speech likelihood exactly. Though it is too slow to compensate a speech recognition system, it enables a more fine-grained assessment of compensation techniques, based on the KL divergence of individual components. This makes it possible to evaluate the impact of approximations that compensation schemes make, such as the form of the mismatch function.
Bibliographic reference. Dalen, Rogier C. van / Gales, Mark J. F. (2010): "Asymptotically exact noise-corrupted speech likelihoods", In INTERSPEECH-2010, 709-712.