<html>
<head>
<meta http-equiv="content-type" content="text/html; charset=ISO-8859-1">
</head>
<body bgcolor="#FFFFFF" text="#000000">
(Apologies for any cross-posting - Please forward to anyone that may
be interested)<b class="moz-txt-star"><span class="moz-txt-tag"><br>
<u><br>
</u><u>POSTDOCTORAL POSITION </u><br>
<br>
</span></b><b class="moz-txt-star"><span class="moz-txt-tag">*</span>SUBJECT<span
class="moz-txt-tag">*</span></b>: Deep neural networks for
source separation and noise-robust ASR
<br>
<b class="moz-txt-star"><span class="moz-txt-tag">*</span>LAB<span
class="moz-txt-tag">*</span></b>: PAROLE team, Inria Nancy,
France
<br>
<b class="moz-txt-star"><span class="moz-txt-tag">*</span>SUPERVISORS<span
class="moz-txt-tag">*</span></b>: Antoine Liutkus (<a
class="moz-txt-link-abbreviated"
href="mailto:antoine.liutkus@inria.fr">antoine.liutkus@inria.fr</a>)
and Emmanuel Vincent (<a class="moz-txt-link-abbreviated"
href="mailto:emmanuel.vincent@inria.fr">emmanuel.vincent@inria.fr</a>)
<br>
<b class="moz-txt-star"><span class="moz-txt-tag">*</span>START<span
class="moz-txt-tag">*</span></b>: between November 2014 and
January 2015
<br>
<b class="moz-txt-star"><span class="moz-txt-tag">*</span>DURATION<span
class="moz-txt-tag">*</span></b>: 12 to 16 months
<br>
<b class="moz-txt-star"><span class="moz-txt-tag">*</span>TO APPLY<span
class="moz-txt-tag">*</span></b>: apply online before June 10 at
<a class="moz-txt-link-freetext"
href="http://www.inria.fr/en/institute/recruitment/offers/post-doctoral-research-fellowships/post-doctoral-research-fellowships/campaign-2014/%28view%29/details.html?nPostingTargetID=13790">http://www.inria.fr/en/institute/recruitment/offers/post-doctoral-research-fellowships/post-doctoral-research-fellowships/campaign-2014/%28view%29/details.html?nPostingTargetID=13790</a>
(earlier application is preferred)
<br>
<br>
Inria is the biggest European public research institute dedicated to
computer science. The PAROLE team in INRIA Nancy, France, gathers
20+ speech scientists with a growing focus on speech enhancement and
noise-robust speech recognition exemplified by the organization of
the CHiME Challenge [1] and ISCA's Robust Speech Processing SIG [2].<br>
<br>
The boom of speech interfaces for handheld devices requires
automatic speech recognition (ASR) system to deal with a wide
variety of acoustic conditions. Recent research has shown that Deep
Neural Networks (DNNs) are very promising for this purpose. Most
approaches now focus on clean, single-source conditions [3]. Despite
a few attempts to employ DNNs for source separation [4,5,6],
conventional source separation techniques such as [7] still
outperform DNNs in real-world conditions involving multiple noise
sources [8]. The proposed postdoctoral position aims to overcome
this gap by incorporating the benefits of conventional source
separation techniques into DNNs. This includes for instance the
ability to exploit multichannel data and different characteristics
for separation and for ASR. Performance will be assessed over
readily available real-world noisy speech corpora such as CHiME [1].<br>
<br>
Prospective candidates should have defended a PhD in 2013 or defend
a PhD in 2014 in the area of speech processing, machine learning,
signal processing or applied statistics. Proficient programming in
Matlab, Python or C++ is necessary. Practice of DNN/ASR software
(Theano, Kaldi) would be an asset.
<br>
<br>
[1] <a class="moz-txt-link-freetext"
href="http://spandh.dcs.shef.ac.uk/chime_challenge/">http://spandh.dcs.shef.ac.uk/chime_challenge/</a>
<br>
<br>
[2] <a class="moz-txt-link-freetext"
href="https://wiki.inria.fr/rosp/">https://wiki.inria.fr/rosp/</a>
<br>
<br>
[3] G. Hinton, L. Deng, D. Yu, G. Dahl, A.-R. Mohamed, N. Jaitly, A.
Senior, V. Vanhoucke, P. Nguyen, T. Sainath, and B. Kingsbury, "Deep
neural networks for acoustic modeling in speech recognition", IEEE
Signal Processing Magazine, 2012.<br>
<br>
[4] S.J. Rennie, P. Fousek, and P.L. Dognin, "Factorial Hidden
Restricted Boltzmann Machines for noise robust speech recognition",
in Proc. ICASSP, 2012.<br>
<br>
[5] A.L. Maas, T.M. O’Neil, A.Y. Hannun, and A.Y. Ng, "Recurrent
neural network feature enhancement: The 2nd CHiME Challenge", in
Proc. CHiME, 2013.<br>
<br>
[6] Y. Wang and D. Wang. "Towards scaling up classification-based
speech separation”, IEEE Transactions on Audio, Speech and Language
Processing, 2013.<br>
<br>
[7] A. Ozerov, E. Vincent, and F. Bimbot, "A general flexible
framework for the handling of prior information in audio source
separation", IEEE Transactions on Audio, Speech and Language
Processing, 2012.<br>
<br>
[8] J. Barker, E. Vincent, N. Ma, H. Christensen, and P. Green, "The
PASCAL CHiME Speech Separation and Recognition Challenge", Computer
Speech and Language, 2013.
</body>
</html>