TPAMI 2012 - Special Issue on Learning Deep Architectures

Special Issue on Learning Deep Architectures
To be published in IEEE Transactions on Pattern Analysis and Machine
Intelligence
http://www.computer.org/cms/Computer.org/transacti...
Topic Description:
In recent years, there has been an emerging interest in architectures,
algorithms, and signal/information processing techniques that learn to
transform data through multiple layers of nonlinearities, hence the
concept
of deep architectures. Several approaches have been developed in the
context
of learning deep architectures, including unsupervised feature
learning using
deep architectures, sparse coding for deep architectures, deep
Boltzmann
machines, stacked auto-encoders, deep belief networks, deep multilayer
perceptrons, convolutional architectures, recursive compositional
models, and
various versions of hierarchical generative models, all of which have
been
successfully applied to a variety of tasks in computer vision, speech
recognition/understanding, audio processing, natural language
processing,
information retrieval, and robotics. These developments cut across
interests
in many traditional and new machine learning, machine intelligence,
pattern
analysis, and signal/information processing research areas.
This special issue invites paper submissions on the most recent
developments
in learning deep architectures and its relation to unsupervised
feature
learning and hierarchical learning algorithms, theoretical
foundations,
inference and optimization, semi-supervised and transfer learning, and
applications to real-world tasks. We also welcome survey and overview
papers
in these general areas pertaining to learning deep architectures.
Detailed
topics of presentations include but are not limited to:
- deep learning architectures and algorithms
- unsupervised feature learning algorithms with deep architectures
- semi-supervised and transfer learning algorithms with deep
architectures
- inference and optimization relevant to learning deep architectures
- theoretical foundations of unsupervised feature learning with deep
architectures
- theoretical foundations of deep learning
- applications of unsupervised feature learning and supervised
learning with
deep architectures
Paper submission and review:
Papers must be submitted online, selecting the choice that indicates
this
special issue. We will accept original research papers and overview/
survey
papers. Peer reviews will follow the standard IEEE review process.
Priority will be given to the papers with high novelty and originality
for the
research papers, and to the papers with high potential impact for
survey/overview papers. Complete manuscripts of full length are
expected,
following the TPAMI guideline in
http://www.computer.org/portal/web/peerreviewjourn....
In case there is uncertainty as to whether the topic of your potential
paper
may fit well to this special issue, you are welcome to contact the
guest
editors below before you start writing the full-length manuscript.
Submission site: https://mc.manuscriptcentral.com/tpami-cs
Dates:
Submissions open period: until April 1st, 2012
First review results: June 15, 2012
Second review results: August 15, 2012
Final manuscripts due: September 1st, 2012
Publication date: December 1st, 2012
Guest editors:
Samy Bengio (bengio-AT-google.com)
Li Deng (deng-AT-microsoft.com)
Hugo Larochelle (hugo.larochelle-AT-usherbrooke.ca)
Honglak Lee (honglak-AT-eecs.umich.edu)
Ruslan Salakhutdinov (rsalakhu-AT-utstat.toronto.edu)
Max Welling (welling-AT-ics.uci.edu)