Navigation

Home
About
Forums
Bug Database
Download
Screenshots

LiarLiar - An opensource computerized voice stress analysis (CVSA) tool

LiarLiar is a voice stress analysis tool for Linux. Voice-stress analysis, an alternative to the polygraph as a method for lie detection, is already widely used in police and insurance fraud investigations. LiarLiar's main purpose is to detect stress in a person's voice. Higher stress levels can be an indication that the person is not being truthful. LiarLiar uses voice stress analysis techniques to attempt to quantify the amount of stress present in speech. LiarLiar is not a lie detector; however, many people will probably use it as such. Moreover, lie detection using voice stress analysis (VSA) techniques have come under much scrutiny as of late. More recent information and technical details about the software can be found in the forums.

Note: The liarliar website crashed, and we are in the process of rebuilding it. Please bear with us.

Features of LiarLiar include:

• Utilizes GStreamer to support multiple input file types (microphone and/or file input)
• Plugin system to support addtional computerized voice stress analysis (CVSA) plugins
• Able to detect microtremors in the 8-14 Hz range.
• Quantification (between 0-100%) of level of stress in speech based upon microtremor analysis
• Written in C++ and utilizing GTK toolkit
• Future versions will make use of additional speech parameters for better accuracy

Changes in version 0.5.2:

• Fixed Microtremor detection
• Fixed Threading issues
• Added an About box
• Ships with an improved FFT plugin

Contact Project Admin: Gene Ruebsamen [gururise@gmail.com]

Screenshots

Sponsors:
Send Free SMS's | S. California Real Estate | Blast Proxy | MySpace Unblocker | Unbiased Skin Care Reviews