Forensic Speaker Recognition Workshop

Details Title: FSI not CSI: Perspectives in State-of-the-Art Forensic Speaker Recognition
Date: 6-7 December 2007
Location: University of New South Wales, Sydney, Australia
Contact: Philip Rose philip.rose@anu.edu.au and Kym Buckley ky.buckley@uws.edu.au


Introduction

How well can you identify the voice of a criminal? In our technological age, more and more offenders are being recorded while committing crimes, and expert opinion is being increasingly sought in the legal process as to whether two or more recordings are from the same speaker. The way forensic speaker recognition is done in real life departs considerably from its depiction in film and popular TV crime shows like Law and Order and Crime Scene Investigation.

The last ten years have witnessed exciting new developments in Forensic Speaker Recognition, not just from the point of methodology but also from the correct evaluation of evidence. To summarise these developments, the HCSNet Forensic Speaker Recognition Workshop is a major forum bringing together for the first time leading practitioners and experts to present State-of-the-Art Forensic Speaker Recognition from several of its most important perspectives, including examples from real case-work.

We invite participants who are involved in any aspect of forensic speaker recognition, but especially researchers in automatic or traditional speaker recognition methods; the evaluation of FSR evidence, and those with actual casework experience including legal counsel, both defence and prosecution.

A special satellite workshop will be held in conjunction with the FSR workshop and be chaired by Helen Fraser from the University of New England. The satellite workshop, is entitled "Can native speakers identify national origin from spoken language: the case of asylum seekers". Please click here for further information and contact details.

Invited Speakers

We have an exemplary line-up of invited speakers for the workshop, as listed below:

Participation and Format

We welcome participation from a range of interests and perspectives within FSR, as outlined above, from both academia and industry, and we would like to involve as many as possible in discussion. Two forms of presentations are possible - short and longer as follows:
  • SHORT Presentations are papers of 5-10 minutes duration, and should outline a perceived problem in FSR and ideas for its solution.
  • LONGER Presentations are of 20-30 minutes, and contain a strong, justified position statement, discussion issue, or results from current research.

For both types of presentation, please submit a short abstract of up to 300 words to Kym Buckley (ky.buckley@uws.edu.au), by the date indicated below.

Funding

HCSNet is able to provide support for a number of participants traveling from outside Sydney. SUBMISSION OF AN ABSTRACT DESCRIBED ABOVE IS A PREREQUISITE FOR BEING CONSIDERED FOR TRAVEL FUNDING. If not all participants can be covered, travel funding may be allocated based on specified interests, as described in the submitted abstract.

Important Dates

  • Friday 9 August: Call for papers and participation
  • Monday 17 September: Registration Opens
  • Monday 8 October: Deadline for receipt of submissions
  • Monday 15 October: Notification of submission acceptance
  • Friday 26 October: Camera-ready final version 300 word abstracts to be received

Organisers

Philip Rose (Australian National University), philip.rose@anu.edu.au
Chris Davis (University of Western Sydney), chris.davis@uws.edu.au
Denis Burnham (University of Western Sydney), d.burnham@uws.edu.au
Kym Buckley (University of Western Sydney), ky.buckley@uws.edu.au
AttachmentSize
satellite_workshop.pdf26.38 KB