Decoder selection based on cross-entropies
Ponani S. Gopalakrishnan, Dimitri Kanevsky, Arthur J. Nadas, D. Nahamoo, Michael Picheny · 2003
The authors generalize the maximum likelihood and related optimization criteria for training and decoding with a speech recognizer. The generalizations are constructed by considering weighted linear combinations of the logarithms of the likelihoods of words, of acoustics, and of (word, acoustic) pairs. The utility of various patterns of weights are examined.>