05110939

j

05110939 Gong, Y. Haton, J.P. Signal-to-String Conversion Based on High Likelihood Regions Using Embedded Dynamic Programming. IEEE Transactions on Pattern Analysis and Machine Intelligence 13, No.03, 297-302 (1991). 1991 Institute of Electrical and Electronics Engineers (IEEE), Washington, DC EN high likelihood regions embedded dynamic programming signal-to-string conversion search Pascal-like language continuous speech recognition 100-word vocabulary dynamic programming pattern recognition search problems speech recognition

doi:10.1109/34.75518

Summary: A method of signal-to-string conversion based on embedded dynamic programming (DP) which can adapt its search to the variation of the input signal is proposed. The optimizing process is guided by high-valued portions of the likelihood function of symbols composing the string and is solved by two embedded dynamic programming processes. Algorithms in a Pascal-like language relating to the solution are given. When applied to continuous speech recognition on a 100-word vocabulary using the phoneme as the basic recognition unit, the method is shown to achieve a 4% improvement in the recognition rate compared to a classical DP-based method.