Jump to: navigation, search

VCS is an abbreviation of Voice Control System. This application is build on a simple generalization of the longest common subsequence algorith, and currently uses the dsp devicefile in some linux-systems for the sound-recording. Once a words has been found in the recording, the energy graph, frequency graph and length of the word is computed, and then the variation of the longest common subsequence algorithm is used to compare the graphs and thereby compare the recorded word to other recordings and thereby identifying the word. This is used to create a menu-system controled by a microphone, where the menu items execute shell commands when they are invoked.


  • Version 0.2: The application uses PortAudio for the sound recording. This increases the portability not just to all recent linux systems but to all major operating systems, and increases the sound deapth and frequency, and therefore allows for a more precise comparrison.
  • Version 0.3: An improved version of the algorithm will be implemented, which will allow a better comparrison (it obtains a statistically better result) using the same amount of calculations.
  • Version 0.4: The application has been split into a library and a 'thin application' that is a frontend to that library. The Idea is that many applications will be able to use the library to enable speech recognition functionality.
  • Version 0.5: The library has been refactorized to a new and improved object oriented design and API.
  • Version 1.0: The frontend has been reimplemented using SdlWidgets to implement an intuitive gui.


Personal tools