In the course of development of speech technologies, we have also developed several C++ libraries intended for digital signal processing as well as general purpose libraries. We have also developed a software tool for speech database labeling and processing.


SLIB is a simple, efficient, lightweight, easy to use, extend and customize C++ library for signal processing. It can be used for standard DSP operations like FIR, IIR, FFT, but also for somewhat more complicated tasks such as speech recognition, voice recognition, speech synthesis etc. It is suitable for both off-line and on-line applications. Since it is highly efficient it can be used in time critical tasks which require speed and no overhead. The library is independent of operating system or target platform. It offers an easy way to observe the signal in any phase of processing, by using included visual tools. Only those tools are platform dependent and now available for Microsoft Windows. The code is written in open source manner and can be downloaded and freely used.

To download version 1.0.0 (complete source: slib, docmaker, slib script, several examples) click on slib_1.0.0.zip.
To download version 1.0.0 of a slib script .exe file (several examples - for those without a C++ compiler) click on slib_script_1.0.0.zip.
To download the entire documentation in English language in HTML format click on slib_doc.zip.

AN_MISC is a general purpose C++ library containing numerous utility functions and classes. It requires SLIB library for proper functioning.

To download version 1.0.0 (complete source, several examples) click on an_misc_1.0.0.zip.
To download the entire documentation in English language in HTML format click on an_misc_doc.zip.

SPEECH_LABEL is a speech labeling tool for ASR and TTS purposes. Its main application is speech database segmentation and phoneme (or other linguistic unit) labeling, i.e. providing information where every linguistic unit begins and ends. Such information can be then saved in a specified format, for future use in ASR and TTS systems, and their initialization and training. Currently there is only the Windows version, but we are planning other OS ports. The software is developed in Visual C++ 6.0, and uses MFC for user interface purposes. It also uses the signal processing library SLIB.

To download the most recent version of this tool (complete source) click on speech_label.zip. Note: since this utility uses SLIB, you should have either its source or a .lib file. The SLIB source should be placed in the same folder as the SpeechLabel project.
To download the executable version of this tool click on speech_label_bin.zip.
To download the entire documentation in English language in HTML format click on speech_label_doc.zip.