Français
1. Background, applications and modeling
Hidden Markov Models (HMM) are statistical models defined by a set of parameters that are learned from a corpus of training data. They implement a probability density on sequential data.
1.1 Illustration of an MMC application
Figure 1 illustrates the main application of MMCs for handwriting recognition. MMC-based systems use an image of a handwritten word to identify the sequence of characters making it up, which means determining the number of characters, recognizing which characters they are, and their start and end positions (abscissae) in the image.
The Ultimate Scientific and Technical Reference
This article is included in
Mathematics
This offer includes:
Knowledge Base
Updated and enriched with articles validated by our scientific committees
Services
A set of exclusive tools to complement the resources
Practical Path
Operational and didactic, to guarantee the acquisition of transversal skills
Doc & Quiz
Interactive articles with quizzes, for constructive reading
Background, applications and modeling
Markov chains and probabilistic finite-state automata
Bibliography
Website
Wikipedia list of tools developed for speech recognition : http://en.wikipedia.org/wiki/List-of-speech-recognition-software
The Ultimate Scientific and Technical Reference
