scholarly journals Identifying unfamiliar voices: the influence of sample duration and parade size.

2021 ◽  
Author(s):  
Nikolas Pautz ◽  
Harriet M J Smith ◽  
Katrin Mueller-Johnson ◽  
Francis .J. Nolan ◽  
Alice Paver ◽  
...  
Keyword(s):  
Identification Performance ◽  
Sample Duration ◽  
Home Office ◽  
Statistical Probability ◽  
Voice Identification ◽  
Innocent Suspect ◽  
The Voice

Voice identification parades can be unreliable due to the error-prone nature of earwitness responses. Home Office guidelines (2003) recommend that voice parades should consist of nine-voices, each played for 60-seconds. This makes parades resource-consuming to construct. In the present paper we conducted two experiments to see if voice parade procedures could be simplified. In Experiment 1, we investigated if reducing the duration of the voice samples on a nine-voice parade would negatively affect performance. In Experiment 2, we first explored if the same sample duration conditions used in Experiment 1 would lead to different outcomes if a six-voice parade were used. Following this, we investigated if there were any difference in identification performance based solely on whether a nine-voice (Experiment 1) or six-voice (Experiment 2) parade was used. Overall, the results suggest that voice durations can be safely reduced without disrupting listener performance. Performance on target-absent parades – which simulate an innocent suspect being apprehended – were at chance-levels in both parade sizes, but the increased number of foils in the nine-voice parade offers increased protection to an innocent suspect by virtue of statistical probability. Thus, we argue that the Home Office guidelines recommending a parade with nine-voices should be maintained.

Voice Identification with the Aid of Spectrographic Analysis

1973 ◽  
Vol 56 (4) ◽  
pp. 944-946
Author(s):  
Ernest W Nash
Keyword(s):  
Objective Evaluation ◽  
Body Parts ◽  
Spectrographic Analysis ◽  
Visual Examination ◽  
Scientific Reason ◽  
Human Voice ◽  
Voice Identification ◽  
The Voice

Abstract The human voice, as an instrument of crime, is used more often than a weapon and automobile combined. Some crimes are committed by the voice alone; therefore, to be able to identify a speaker by his voice is a very desirable goal in the fight against crime. However, desire has been somewhat hindered by the lack of technology and instrumentation. The use of spectrograms (voiceprints) to assist the expert in making an objective evaluation of the voices in question is discussed. The scientific reason for accepting the identification of a speaker’s voice is the uniqueness of man. Therefore, if a unique person uses unique physiological body parts to produce the sounds of speech, it logically follows that sound will also be unique. By the visual examination of the spectrographic analysis, a trained expert is able to compare the uniqueness.

scholarly journals Pengaruh Perubahan Pitch dan Formant Terhadap Hasil Identifikasi Kepemilikan Suara Dengan Metode Audio Forensik

2021 ◽  
Vol 9 (1) ◽  
pp. 1-8
Author(s):  
Mifta Nur Farid ◽  
Dani Dwi Putra ◽  
Barokatun Hasanah
Keyword(s):  
Analysis Of Variance ◽  
Voice Recognition ◽  
Frequency Characteristics ◽  
Sound Recordings ◽  
Voice Identification ◽  
Recorded Sound ◽  
Audio Forensics ◽  
The Difference ◽  
Pitch Analysis ◽  
The Voice

Audio forensics is a field of science that analyzes audio such as sound recordings. Voice recordings always have information in the form of frequency characteristics, the identities of these frequencies can be identified. Furthermore, an analysis of changes in pitch and formant will be carried out. This study used pitch analysis and analysis of variance on formants. With the correct procedure for handling recorded sound evidence which is then followed by procedural examination and analysis, it is hoped that the results of the voice recognition examination can scientifically show the ownership of the voice in the recording. Based on the results of the overall analysis of the sound recordings of evidence and comparison after carrying out various stages of analysis, the voice recordings are "not identical" from the same person. The thing that causes the inequality in voice identification is the difference in intonation or tone of the subject's speech when the voice is recorded.

scholarly journals A flexible method to create wave file features

2021 ◽  
Vol 11 (2) ◽  
pp. 1311
Author(s):  
Ghazi M. J. Qaryouti
Keyword(s):  
Audio Signal ◽  
Digital Audio ◽  
Human Knowledge ◽  
Signal Identification ◽  
Important Data ◽  
Recognition Time ◽  
Audio File ◽  
Voice Identification ◽  
Small Set ◽  
The Voice

Digital audio signal is one of the most important data type at present, it is used in various vital applications, such as human knowledge, security and banking applications, most applications require signal identification and recognition, and to increase the efficiency of these applications we must seek a method to represent the audio file by a small set of values called a features vector. In this paper research we will introduce an enhanced method of features extraction based on k-mean clustering. The method will be tested and implemented to show how the proposed method can reduce the efforts of voice identification, and can minimize the recognition time a set of voice extracted features must be used instead of using the voice wave file.

scholarly journals VOICE IDENTIFICATION BASED ON THE I-VECTOR AND DEEP NEURAL NETWORKS USING SHORT UTTERANCES

2020 ◽  
Vol 3 (77) ◽  
pp. 82-90
Author(s):  
O. Mamyrbayev, ◽  
◽  
A. Akhmediyarov, ◽  
A. Kydyrbekov, ◽  
N. Mekebayev, ◽  
...  
Keyword(s):  
Deep Neural Networks ◽  
Voice Recognition ◽  
Vector System ◽  
Nonlinear Mapping ◽  
Voice Identification ◽  
Large Spread ◽  
Short Utterances ◽  
User Recognition ◽  
The Voice ◽  
Distinguishing Features

Text-independent voice recognition of the user using short sentences is a very difficult task due to the large spread and inconsistency of the content between short sentences, in order to improve user recognition by voice, it is planned to highlight several sets of distinguishing features that contain more information related to the voice. The results show that the i-vector DNN system is superior to the GMM i-vector system for various durations. However, the characteristics of both systems deteriorate significantly as the duration of the sentences decreases. To solve this problem, we propose two new nonlinear mapping methods that train DNN models to map i-vectors extracted from short sentences to their corresponding i-vectors of long sentences.

Performance of VoIP in Wired-Cum-Wireless Ethernet Network

2012 ◽  
Vol 4 (4) ◽  
pp. 1-25
Author(s):  
Nurul I. Sarkar ◽  
Kashif Nisar
Keyword(s):  
System Performance ◽  
Low Cost ◽  
Local Area ◽  
Data Networks ◽  
Data Traffic ◽  
Home Office ◽  
Traffic Characteristics ◽  
Packet Arrival ◽  
Campus Networks ◽  
The Voice

The Voice over Internet Protocol (VoIP) is a rapidly growing technology that enables transport of voice over data networks such as Ethernet Local Area Networks (LANs). This growth is due to the integration of voice and data traffic over the existing network infrastructure, low cost, and improved network management offered by the technology. This paper reports on the performance of VoIP traffic characteristics in a wired-cum-wireless Ethernet LAN. The effect of increasing the number of VoIP wireless clients, different voice codec schemes, and packet arrival distributions on system performance is investigated. Through various simulation experiments under realistic network scenarios, such as Small Office Home Office (SOHO) and campus networks, this paper provides an insight into the performance of VoIP over Ethernet LANs. Simulation results show that VoIP clients and voice codec schemes have significant effect on system performance. The authors preformed OPNET-based simulations to validate their experiments.

scholarly journals THE IMPORTANCE OF VOICE IDENTIFICATION IN THE WITNESS RECOGNITION PROCEDURE

TEME ◽  
2020 ◽  
pp. 1157
Author(s):  
Jadranka R Otašević ◽  
Saša Atanasov
Keyword(s):  
Voice Recognition ◽  
Legal Framework ◽  
Point Of View ◽  
Theoretical Point ◽  
Witness Testimony ◽  
Voice Identification ◽  
Procedure Code ◽  
The Republic ◽  
Direct Recognition ◽  
The Voice

From a theoretical point of view, this paper considers the evidentiary action of recognizing the voice of the perpetrator by the witness. It is the identification of the voice by a person who is usually an "unprofessional listener". Due to the specificity of the voice as an object of recognition, the involvement of forensics (linguists and phoneticians) in the organization and immediate realization of the voice recognition action seems inevitable. Their activity would be manifested in giving guidance to the authority on how to increase the efficiency of voice identification and the accuracy of witness testimony. The witness gives evidence based on his perceptual (auditory) abilities in a procedure prescribed by the law, in which the credibility of his/her testimony is simultaneously checked and assessed. The Criminal Procedure Code of the Republic of Serbia establishes the legal framework for taking the voice recognition action, while the content of performing the direct recognition action is determined by the criminal-tactical rules.

Driver Voice Identification System Using Auto-Correlation Function and Average Magnitude Difference Function

2014 ◽  
Vol 490-491 ◽  
pp. 1287-1292 ◽  
Author(s):  
Jian Da Wu ◽  
Pang Yi Liu ◽  
Guan Long Hong
Keyword(s):  
Correlation Function ◽  
Identification System ◽  
Zero Crossing ◽  
Voice Identification ◽  
Auto Correlation ◽  
Difference Function ◽  
Auto Correlation Function ◽  
The Voice ◽  
Magnitude Difference ◽  
Average Magnitude Difference Function

This study presents a driver identification system using voice analysis for a vehicle security system. The structure of the proposed system has three parts. The first procedure is speech pre-processing, the second is feature extraction of sound signals, and the third is classification of driver voice. Initially, a database of sound signals for several drivers was established. The volume and zero-crossing rate (ZCR) of sound are used to detect the voice end-point in order to reduce data computation. Then the Auto-correlation Function (ACF) and Average Magnitude Difference Function (AMDF) methods are applied to retrieve the voice pitch features. Finally these features are used to identify the drivers by a General Regression Neural Network (GRNN). The experimental results show that the development of this voice identification system can use fewer feature vectors of pitch to obtain a good recognition rate.

Sign in / Sign up

Export Citation Format

Share Document