[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: best Octave developers platform?

From: Adrian Marks
Subject: Re: best Octave developers platform?
Date: Thu, 11 Jan 2007 14:26:08 -0500
User-agent: Opera M2/8.51 (Win32, build 7712)

   Dear Octave List,

These annoucements came to me from another list but perhaps some of you here would be interested in participating? I could be a good opportunity to show Octave to the world stage of research developement.


+++++++++++++++++++ PART 1 (Of 2 Parts) ++++++++++++++++++++++++++++++


                 1st Stereo Audio Source Separation
                        Evaluation Campaign


               Participation deadline: April 13, 2007

Do you have an algorithm for source separation from few channels? Or are you interested in this challenging issue? Then you should take part in the first
campaign for the evaluation of stereo audio source separation algorithms.

The task is to estimate from a two-channel speech or music mixture the
contribution of each source on each channel, the number of sources being equal to three or more. This campaign complements recent evaluation initiatives conducted by Lucas Parra and the PASCAL network, which focused on different numbers of
sources and channels.

Three types of mixtures are considered:
- instantaneous mixtures,
- synthetic convolutive mixtures,
- live recordings.

All contributions are welcome, including established, novel, blind or non-blind, two-channel or single-channel algorithms. The evaluation will be non-competitive: the results will be made available on the campaign website for listening and evaluated using multiple criteria, possibly proposed by the participants. The results will be summarized in a paper to be discussed at the ICA'07 conference (see, during which the participants who wish to present a poster about their algorithm will also have the opportunity to do so. These
posters will then be published on the campaign website.

For more information about the campaign and to download training/test sets, see

Best regards,

Emmanuel Vincent, Hiroshi Sawada, Pau Bofill, Shoji Makino and Justinian Rosca

+++++++++++++++ PART 2 ++++++++++++++++++++++++++++++++

Dear List

Speech Separation Challenge 2!

As a follow on to the highly successful PASCAL Speech Separation Challenge (http:/ / we are pleased to
announce the Speech Separation Challenge Part 2! The new task addresses the
problem of separating and recognising overlapping speech from multi-channel
recordings made in realistic environments. Results will be presented at a special session of MLMI07 ( taking place in Brno
in the Czech Republic from 28-30 June 2007.

The task is to separate the speech of 2 overlapping speakers recorded in a real
meeting room using two eight element microphone arrays, and subsequently
recognise each speaker. Large quantities of development data are provided. The speakers read Wall Street Journal sentences, and a standard WSJ recogniser is
provided for recognition of the separated speech, although participants are
welcome to use their own recognition systems if they wish.

Contributions are welcomed from a variety of different approaches including well- known existing algorithms such as blind source separation and beamforming, as
well as novel techniques.

Full details of the challenge plus downloads for the data and default recognition
system are available from the challenge website :

* Submission deadline for MLMI07 papers for the SSC2 special session is May 1st
2007 (note, this is later than the standard MLMI paper submission deadline)

Best regards

Mike Lincoln


Mike Lincoln
Centre for Speech Technology Research
University of Edinburgh

reply via email to

[Prev in Thread] Current Thread [Next in Thread]