Skip to content Skip to main navigation Skip to footer

Key Features (VIN)

Phonexia Voice Inspector software offers several features that strongly support the work of voice forensic experts:

  • A standalone application with a complete easy-to-use Graphical User Interface (GUI) 
  • Automatic comparison of questioned recording (unknown speaker recording or voiceprint) against a suspected reference speaker (group of recordings or voiceprints) with a known speaker i.e. 1:1 identification and 1:N identification.
  • Implemented speech technologies: Speaker Identification, Speaker Diarization, Phoneme Recognizer, Voice Activity Detection, Speech Quality Estimation
  • A search for repetitive sound patterns across all recordings in audio due to the automatic phonemic transcription
  • An inbuilt tool that helps transcribe selected parts of recordings and skip other parts
  • Input:
    • Questioned recordings (a minimum of 1 recording, a minimum of 7 seconds of net speech in the recording) 
    • Suspected speaker recordings (a minimum of either 3 recordings with at least 7 seconds of net speech in each, or 1 recording with at least 20 seconds of net speech)
    • The Population set (a technical minimum of 10 speakers, a recommended minimum of 30–40 speakers. Depending on the type of score calculation, each speaker should have one or more recordings with at least 7 seconds of net speech per each recording.)
  • Supported audio format: MS Wave or RAW with linear coding (8 or 16 bits), A-law, Mu-law; Sampling frequency 8kHz or higher
  • Compatibility with the widest range of audio sources possible (applies channel compensation techniques verified by NIST evaluation): GSM/CDMA, 3G, VoIP, landlines, etc.
  • Output: 
    • A scoring table with the results of comparisons in a Likelihood Ratio, Log-Likelihood Ratio (decimal or natural logarithm), and Verbal Ratio
    • The graphical presentation of results in the form of a Probability Density Function plot and a Tippet plot
    • The Diarization panel with labels for different speakers
    • The Phoneme transcription panel for discovering similar sound patterns across recordings
    • The Voice Activity Detection panel with labels for speech and non-speech segments
    • The Spectrum panel and the Spectrum details layout
    • An editable report with all results (including the scoring table and graphs) in one document (exportable as PDF or OpenDocument)
  • Available platforms: Windows 64bit (x86_64) or Linux 64bit (x86_64)