Voice Print Analysis

Voice Print Analysis, VPA at Sunartek, is an advanced informative speech analysis system, capable of analyzing audio files for Speech/Non-speech detection, Language identification, and Speaker identification. Sunartek ‘s VPA is based on Client/server architecture.

Voice Print Analysis or Voice identifier systems are mainly deployed by the defense and investigation agencies.

Features of Voice Print Analysis

Speech Quality Estimator

Speech Transcription

Processing Speed

Features

Speech Quality Estimator

Speech Quality estimator examines the quality of the voice sample and determines if the voice sample can be processed through language detection, speaker identification, etc, based on various quality analytical parameters.

Speech Transcription

Speech Transcription accurately transforms voice samples into plain text automatically to uncover the content hidden in the voice sample and to detect the emerging trends. This analysis is very useful for defense and investigation agencies.

Processing Speed

The processing speed of Sunartek’s VPA is approximately 7x faster than real-time processing on 1 CPU core.

Functionality

speech detection

Speech Detection

Speech detection features marks and extracts the areas in the sample which contain speech by understanding the threshold values and frequency bands.

Speaker Identification

Speaker Identification feature identifies the speaker instantly with high accuracy by comparing the speech sample with the pre-trained speech dataset.

Speaker Identification

Speaker Identification feature identifies the speaker instantly with high accuracy by comparing the speech sample with the pre-trained speech dataset.
Gender Identification

Gender Identification

Gender Identification feature recognizes whether the voice belongs to a male or a female based on the voice frequency.

Language Identification

Language identification feature analysis the language of the speaker by comparing it with the acoustic characteristics of each language and providing an identification score.
Language Identification
Language Identification

Language Identification

Language identification feature analysis the language of the speaker by comparing it with the acoustic characteristics of each language and providing an identification score.
Age Identification

Age Detection

Sunartek’s VPA estimates the age results of the speaker individually or even in a group.

Emotion Detection

Emotion detection feature assesses the mood of the speaker by analyzing the conversation.
Emotion Identification
Emotion Identification

Emotion Detection

Emotion detection feature assesses the mood of the speaker by analyzing the conversation.

Keyword Spotter

Keyword Spotting feature automatically identifies the occurrences of specific keywords and keyphrases in the voice sample.
Input format in Sunartek's Voice Print Analysis

WAV or RAW, FLAC, OPUS

Output format in Sunartek's Voice Print Analysis

Result files are in XML/JSON format.

Languages supported in Sunartek's VPA

Supported Languages in Sunartek’s VPA for speech transcription are Arabic Gulf, Arabic Levantine, Chinese, Croatian, Czech, Dutch, English US, Farsi, French, German, Italian, Pashto, Polish, Russian, Slovak, Spanish, Swedish. New languages can also be trained.

Applications of Sunartek's Voice identifier for defense and investigation

The Voice Identifier system determines the keywords present in the voice sample, which will be very helpful for investigation.   If the database of the investigative agency has voice samples, it also helps to determine the speaker. Overall voice print analysis system simplifies the process of investigation.

Other Solutions