SVID SPEAKER VOICE IDENTIFICATION SOLUTION

ACU-EXPERT SVID solution is specially designed for the purposes of forensic speaker voice identification including tools for all contemporary approaches to the task:

• Automatic speaker recognition technologies
• Extraction and comparison of linguistic features
• Detailed speech analysis based on acoustic speech production theory

The solution consists of two parts: ACU-EXPERT option NF for automatic comparison based on voice biometrics and automatic speaker gender and language identification and ACU-EXPERT option OT – software application for accurate speech marking, text decoding, voice and speech features automated extraction and comparison.

SVID solution covers full range of existing methods of forensic speaker voice identification and allows to guarantee experts' confidence and conclusions.

FEATURES

EXTENSIVE FORMAT SUPPORT - All audio and majority of video file formats supported (with simultaneous video playback), precise audio re sampling.

AUTOMATIC SEGMENTATION of speakers in two or three persons dialogue to extract one speaker from the flow.

SPEECH PRE-PROCESSING including audio normalisation, noise reduction and speech enhancement. Channel adaptation, channel difference compensation to avoid cross-channel influence.

TEXT DECODING AND PHRASE MARKING for conduction of by-word content of the conversation.

SPEAKER DATABASE WITH CUSTOMIZED DATABASE RECORD PROFILE for proper reference population collection and audio storing. 68 built in language profiles and functionality to add new language/dialect.

ASSESSMENT OF SUITABILITY for automatic comparison and expert investigation with speech quality parameters export.

AUTOMATIC PITCH EXTRACTION with pitch distribution comparison and speech melodic patterns comparison.

FORMANT EXTRACTION and analysis of formants statistics in vowels.

IN-BUILD IPA CHART

AUTOMATED SEARCH of matching words and phonemes for detailed comparison.

REPORT PREPARATION for court testimony.