Multiplatform server-based speech recogniser generator for embedded systems

Automatic speech recognisers still more come in the practice of regular users of computers and various devices. However, so far there is no speech recogniser on the market in the form of an electronic component or a module that could be easily usable in embedded systems. Our intention is to prepare a theoretical concept and practical realisation of a software tool that facilitates the development of the standalone module of spoken commands recogniser and enables the easy configurability during a possible purchase in an e-shop for each individual part. The first step towards this goal is our phpHMM software tool that facilitates the development and optimisation of speech recogniser algorithms. We use it for generating source code of speech recogniser using the PHP scripting language and MySQL database. The input of the system is a speech model and a list of words to be recognised. The output is the source code and data structures in the C programming language, which are consecutively compiled into an executable program.