Research Projects


Making demonstrably valid and reliable forensic voice comparison a practical everyday reality in Australia

Australian Research Council Linkage Project (LP100200142), 2010–2013

Participant Information Statement and Consent Form

Brief Description:

To make forensic voice comparison a practical everyday reality in Australia, for use in police investigations and for presentation of evidence in court, forensic scientists must have a practical and demonstrably reliable forensic-voice-comparison system including a large representative database of Australian English voices. In collaboration with Australian police forensic laboratories and other partners we will develop and test such a system, improving on existing systems by combining the strengths of acoustic-phonetic and automatic approaches. The system will comply with the strictest international standards for the admissibility of scientific evidence in court, using the same evaluative framework as is applied to DNA.

Long Description (from funding application)

Principal Investigator:

Other Investigators:

Partner Organisations:

Consultants:

Funding:

  • AU$ 278 000 from ARC
  • AU$ 100 000 cash from partners
  • AU$ 166 013 in-kind from partners
  • AU$ 544 013 total value

43% of all project proposals were funded this round. The average funding for successful proposals this round was 68% of the amount requested. We received 91% of our requested funding amount.




Incorporation of forensic analysis techniques as part of an automatic speaker recognition system

United States Government, Office of the Director of National Intelligence (ONDI), Intelligence Advanced Research Projects Activity (IARPA), through the Army Research Laboratory (ARL), Broad Agency Announcement IARPA-BAA-09-02, Biometrics Exploitation Science and Technology (BEST)

Subcontract through IBM Thomas J. Watson Research Center, 2010–2011

Brief Description:

The overall scope of the project is to examine the incorporation of forensic analysis techniques as part of an automatic speaker recognition system. The UNSW team’s role in the work is to examine phonetic attributes that (i) differentiate between speakers and (ii) that could be considered for use in an automatic speaker recognition system. The UNSW team will also examine problematic audio utterance pairs to observe and identify any potential information that could be exploited. An aggressive challenge for the UNSW team is to perform research into identifying speaker attributes that may be consistent for languages other than English and across languages.

Principal Investigator at IBM:

Other IBM researches:

Principal Investigator at UNSW:

Other UNSW researchers and subcontractors: