Personal tools

Blizzard Challenge 2011 Workshop

From SynSIG
The Blizzard Challenge 2011 workshop is generously hosted by Loquendo http://www.loquendo.com


Friday 2nd September 2011

The workshop will be hosted by Loquendo at their premises in Turin, Italy. Turin is about 3 hours from Florence by train.


Contents

Call for participation

The Blizzard Challenge 2011 Workshop is the culmination of the Blizzard Challenge 2011 which is an open speech synthesis evaluation campaign using common data sets and a large listening test. The aims of the workshop are to present the results from the listening tests and for participants in the Challenge to describe their systems. This year we also invite people to showcase their preliminary systems for the 2012 Challenge, using audiobook data.

The workshop is a satellite of Interspeech 2011

Who can attend the workshop ?

The workshop is open to all and we encourage participation from anyone interested in speech synthesis.


Who can submit a paper to the workshop ?

All participants in the Challenge are expected to submit a paper describing their entry (even if they cannot attend the workshop in person). Papers will be refereed by the Programme Committee.

Programme Committee

  • Simon King, University of Edinburgh, UK
  • Alan Black, Carnegie Mellon Univerisity, USA
  • Keiichi Tokuda, Nagoya Institute of Technology, Japan

Paper submission instructions

  • Use the Interspeech 2011 authors' kit, but your paper should be up to SIX pages in length
  • Remember that Blizzard is a scientific investigation - we are all trying to understand why some techniques work better than others.
  • With this in mind, please write a detailed, technical paper aimed at a specialist audience. Focus on analysis and evaluation. Try to explain WHY your system performed the way it did, and what makes it different from other systems. Explain why your system is designed in a particular way. For example, report internal evaluations you have done to select certain methods.
  • Submit your paper by email to blizzard@festvox.org by the 31st July 2011 and we will notify you of acceptance by 5th August 2011. If you need an earlier notification (e.g., in order to obtain travel permission from your employer), then simply submit the paper earlier and let us know that you need a fast decision on acceptance.

Location and date

Date: Friday 2nd September 2011 (all day)

Host: Loquendo, Turin, Italy.

Venue (note - not Loquendo's premises):

Telecom Italia

Largo Borgaro, 27

10149 Torino

Italy

Map

Street view

Programme

The workshop format is single-track with oral presentations from the participants in the Challenge, the organisers and invited speakers.

Each system presentation should last for a maximum of 15 minutes, including time for questions. Presenters should bear in mind that the audience will mainly comprise speech synthesis experts; therefore presentations do not need to include extensive background material.

We will start at 9am and run until around 6pm. The programme is as follows:

  • 08.30 Registration desk opens
  • 09.00-10.00 Welcome, introduction, summary
    • Simon King, on behalf of the organisers
    • Welcome message from Loquendo
    • Message from Google
  • 10.00-11.00 System presentations
    • Lessac Technologies
    • Institute for Language and Speech Processing
    • Nagoya Institute of Technology
  • 11.00-11.30 Coffee break
  • 11.30-12.30 System presentations
    • Aholab (UPV) & TALP (UPC)
    • Politehnica University of Bucharest
    • Helsinki University & Aalto University
  • 12.30-14.00 Lunch
  • 14.00-15.00 System presentations
    • Institute for Infocomm Research (I²R)
    • University College Dublin
    • University of Science and Technology of China
  • 15.00-15.45 Coffee break
  • 15.45-16.15 Evaluation
    • Presentation (20mins): Florian Hinterleitner et al: a proposed evaluation protocol for an audiobook reading task
    • Presentation (15mins): Sajad Shirali-Shahreza: Verifying Human Users in Speech-Based Interactions (a possible future Blizzard Challenge could look at defeating such systems using synthesis)
  • 16.15-onwards Blizzard 2012
    • Informal “show and tell” of preliminary systems built from the 2012 data (~5 mins each): USTC, Lessac Technologies, CSTR (Edinburgh), ILSP, NIT, .... more welcome (tell us in advance, or on the day)!
    • Open discussion: the main topic will be "How to perform the evaluation in the Blizzard Challenge 2012"
  • ~18.00 Close
  • There will probably be an informal social gathering after the workshop for drink/food - details to follow.

Practical information

Registration

Registration procedure

Each person registering for the workshop must send the following information, to meet Loquendo's security requirements:

  • Family name
  • Given name
  • Affiliation
  • Passport number

Please mail blizzard@festvox.org by 31st July 2011 to register. In your email, please use the subject line "Registration for the Blizzard Challenge Workshop 2011". Please send one email per person registering.

All registrations will be acknowledged. If the workshop is oversubscribed, priority will be given to participants in the 2011 Challenge and participants in the 2012 Challenge.

Note that this year you must register for the workshop before receiving notification that your paper has been accepted. If this causes you any problems (e.g. in gaining permission to attend from your employer), please contact blizzard@festvox.org for assistance.

  • If you are the presenter of a paper, please mention that in your registration email
  • If you wish to present your preliminary system for the 2012 Challenge, please mention that too. The format of these presentations will be decided later, but please plan on playing some example output.

Cost

There is no cost to attend the Blizzard Workshop this year, thanks to support from Loquendo.

Accommodation and travel

Directions to Loquendo and suggestions of hotels are provided in this document

Published proceedings

The papers will be published on festvox.org


SynSIG is a Special Interest Group of ISCA, the International Speech Communication Association.

SynSIG 1998-2019