In this article we present the latest version of VOCALISE (Voice Comparison and Analysis of the Likelihood of Speech Evidence), a forensic automatic system for speaker recognition. VOCALISE, with selectable state-of-the-art and legacy speaker modelling algorithms allows the forensic practitioner to work with spectral features (such as Mel Frequency Cepstral Coefficients (MFCCs)), phonetic features (such as formants), or features of their own choice (such as voice quality metrics, articulation rate, etc.). It is capable of comparing features from a test audio file of a target speaker against features from an audio file of a suspected speaker, or an entire list of suspected speakers, and produces a likelihood score or likelihood ratio for each comparison. It is built with an ‘open-box’ architecture that transparently allows the user to provide their own data to train the system’s algorithms. These algorithms include Gaussian Mixture Modelling (GMM) with (and without) MAP (maximum a posteriori) adaptation, i-vector extraction with PLDA (Probabilistic Linear Discriminant Analysis) and cosine distance comparison. VOCALISE seeks to form a bridge between traditional forensic phonetics-based speaker recognition and forensic automatic speaker recognition.
Cite as: Alexander, A., Forth, O., Atreya, A. A., Kelly F. (2016) VOCALISE: A forensic automatic speaker recognition system supporting spectral, phonetic, and user-provided features. Odyssey 2016 Show & Tell, http://www.odyssey2016.org/papers/Show_tell/88.pdf.
@Misc{Alexander+2016,
author = {Anil Alexander and Oscar Forth and Alankar Aryal Atreya and Finnian Kelly},
title = {VOCALISE: A forensic automatic speaker recognition system supporting spectral, phonetic, and user-provided features},
booktitle = {Odyssey 2016: The Speaker and Language Recognition Workshop, Show And Tell },
address = {Bilbao, Spain},
year = {2016},
month = {June 21-24},
url = {http://www.odyssey2016.org/papers/Show_tell/88.pdf}
}
The aim of the described system is to provide an online solution that profiles customers of a call centre. As an auxiliary module it might enhance functionality of modern call centre systems by active voice analysis. Integrated with existing databases, our system allows for analysis of constant and temporal caller characteristics during a call — respectively identity, age, gender, emotional state, speech rate and an acoustic background. The specifically developed tool both shortens call time and enhances the amount of information gathered, and consequently reduces cost and workload of call centre responders. Index Terms: speaker recognition, emotion detection, age detection, gender detection, acoustic background detection, call centre support, voice biometrics.
Cite as: Witkowski, M., Gałka, J., Grzybowska, J., Igras, M., Jaciów, P., Ziółko, M. (2016) Online caller profiling solution for a call centre. Odyssey 2016 Show & Tell, http://www.odyssey2016.org/papers/Show_tell/89.pdf.
@Misc{Witkowski+2016,
author = {Marcin Witkowski and Jakub Gałka and Joanna Grzybowska and Magdalena Igras and Paweł Jaciów and Mariusz Ziółko},
title = {Online caller profiling solution for a call centre},
booktitle = {Odyssey 2016: The Speaker and Language Recognition Workshop, Show And Tell },
address = {Bilbao, Spain},
year = {2016},
month = {June 21-24},
url = {http://www.odyssey2016.org/papers/Show_tell/89.pdf}
}