ESMERALDA is a development environment for statistical
recognizers operating on sequential data (speech, handwriting,
biological sequences). It supports continuous density Hidden
Markov models, Markov chain (N-gramm) models, and Gaussian
mixture models.