UNSW Forensic Speech Science Conference 2012

Forensic Voice Comparison LaboratoryAustralasian Speech Science and Technology Association (ASSTA)

Forensic Voice Comparison Laboratory
School of Electrical Engineering & Telecommunications
University of New South Wales

Sydney, Australia

Satellite event to the 14th Australasian International Conference on Speech Science and Technology.

last update: 2012-12-05

Introduction

The idea of this conference is to provide a venue for presentations which would not be presented in the main SST-12 conference. For example, preliminary work for which no paper has yet been written, or work for which refereed papers have already been submitted to other venues. This is intended to be a relatively informal conference.

Important dates

* Maximum capacity is 70. On-site registration will continue until maximum capacity is reached.

Registration

Registration is free to registrants of the SST-2012 conference. Registration for others is AUD 20 (Students AUD 10), payable on the day of the conference.

To register, send an e-mail to registration e-mail address with the following information:

Lunch and dinner are not included, but locations of local restaurants will be announced on the day of the conference.

Conference programme

9:00 - 9:25 Registration
9:25 - 9:30 Introduction
  Database selection & size
9:30 - 9:50 Morrison, Ochoa Database selection for forensic voice comparison
9:50 - 10:10 Rose The Likelihood Ratio goes to Monte Carlo - the effect of reference sample size on likelihood-ratio estimates
  Articulation-rate based FVC
10:10 - 10:30 Gold Articulation rate as a discriminant in forensic speaker comparisons
10:30 - 11:00 break
  Perception-based study
11:00 - 11:20 Didla Pitch breaks as a voice disguise
  Simulating recording conditions
11:20 - 11:40 Ochoa Simulating casework recording conditions for forensic voice comparison
11:40 - 12:00 Alzqhoul, Nair, Guillemin Platforms for simulating the impact of mobile phone networks on speech
12:00 - 1:30 Lunch break
  PCA kernel density approach
1:30 - 1:50 Nair, Alzqhoul, Guillemin A new approach to computing likelihood ratios based on principal component analysis
1:50 - 2:10 Lu, Liu, Guillemin Comparative performance of real cepstral coefficients vs mel frequency cepstral coefficients for forensic voice comparison
2:10 - 2:30 Nair, Alzqhoul, Guillemin Impact on forensic voice comparison of speech acquired from the GSM mobile phone network – some preliminary results
2:30 - 3:00 break
  Formant-based FVC
3:00 - 3:20 Susanto Synchronic Stability Vowel System in Bahasa Indonesia forensic speaker identification
No cancellation notification was received, but this speaker did not attend the conference, and the talk was not presented.
3:20 - 3:40 Pingjai Forensic Voice Comparison in Standard Thai using vowel formants
3:40 - 4:00 Zhang, Enzinger, Morrison, Ochoa Effects of telephone transmission on the validity of formant-trajectory-based forensic-voice-comparison systems
  Modeling issue
4:00 - 4:20 Morrison, Ochoa, Lindh Calculating the reliability of likelihood ratios: Addressing modelling problems related to small n and tails
4:20 - 4:30 Conference closing

Venue & travel information

Getting to UNSW Kensington campus

Buses: NSW Transport Info

Taxi/cab:

Google Maps view of UNSW Kensington campus

UNSW Kensington campus

The conference is held in Room G3 located on the ground floor in the Electrical Engineering & Telecommunications building (G17). Note: Entry to the EE&T building is on the lower ground (LG) floor. Use the lift or the stairway to get to the ground floor.

UNSW Kensington campus map

Accommodation

Some local accommodation options (we do not endorse any particular establishment).

Organizers:

Ewald Enzinger and Felipe Ochoa

Contact us at contact e-mail address.