Rebecca Martin | Knowledge 4 All Foundation Ltd.

March 12, 2013

PASCAL2

Researcher position, Statistical Machine Translation – Xerox Grenoble, France

Researcher, Statistical Machine Translation

The Machine Learning for Document Access and Translation group of the Xerox Research Centre Europe conducts research in Statistical Machine Translation and Information Retrieval, Categorization and Clustering using advanced machine learning methods.

We are opening a position for a researcher with a background in Statistical Machine Translation to support our participation in the EU-funded project TransLectures (www.translectures.eu).

Required experience and qualifications:

– PhD in computer science or computational linguistics, with focus on
statistical methods
– Experience with Statistical Machine Translation.
– Good publication record and evidence of implementing systems.
– A good command of English, as well as open-mindedness and the will
to collaborate within a team.
– Acquaintance with Spoken Language Translation is a plus.

Preferred starting date: July 2013

Contract duration: 18 months

Application instructions

Please email your CV and covering letter, with message subject “Statistical Machine Translation Researcher” to xrce-candidates and to Nicola.Cancedda at xrce.xerox.com. Inquiries can be sent to Nicola.Cancedda at xrce.xerox.com.

XRCE is a highly innovative place and we strongly encourage publication and interaction with the scientific community.

Job announcement URL:

http://www-int.xrce.xerox.com/About-XRCE/Career-opportunities/Researcher-Statistical-Machine-Translation

March 11, 2013

PASCAL2

Second Call for papers SISAP 2013

===========================
SISAP 2013 – SECOND CALL FOR PAPERS
===========================

We apologize if you receive duplicates of this CFP.
Please feel free to distribute it to those who might be interested.

SISAP 2013: 6th International Conference on Similarity Search and Applications
October 2-4, 2013
A Coruna, Spain

Web site: http://www.sisap.org/2013
Facebook: http://www.facebook.com/sisap2013
Twitter: http://twitter.com/sisap2013,

Scope

The International Conference on Similarity Search and Applications (SISAP) is an annual forum for researchers and application developers in the area of similarity data management. It aims at the technological problems shared by numerous application domains, such as data mining, information retrieval, computer vision, pattern recognition, computational biology, geography, biometrics, machine learning, and many others that need similarity searching as a necessary supporting service.

The SISAP initiative (www.sisap.org) aims to become a forum to exchange real-world, challenging and innovative examples of applications, new indexing techniques, common test-beds and benchmarks, source code and up-to-date literature through its web page, serving the similarity search community. Traditionally, SISAP puts emphasis on the distance-based searching, but in general the conference concerns both the effectiveness and efficiency aspects of any similarity search problem.

The series started in 2008 as a workshop and has developed over the years into an international conference with Lecture Notes in Computer Science (LNCS) proceedings. As in previous editions, a small selection of the best papers presented at the conference will be recommended for inclusion in a special issue of Information Systems. In October 2013, SISAP will take place in A Coruna, Spain.

Keynote Speakers

Ricardo Baeza-Yates (VP of Yahoo! Research for Europe, Middle East and Latin America)
Jiri Matas (Center for Machine Perception, Czech Technical University)

Topics of interest

The specific topics include, but are not limited to:
Similarity queries – k-NN, range, reverse NN, top-k, etc.
Similarity operations – joins, ranking, classification, categorization, filtering, etc.
Evaluation techniques for similarity queries and operations
Merging/combining multiple similarity modalities
Cost models and analysis for similarity data processing
Scalability issues and high-performance similarity data management
Feature extraction for similarity-based data findability
Test collections and benchmarks
Performance studies, benchmarks, and comparisons
Similarity Search over outsourced data repositories
Similarity search cloud services
Languages for similarity databases
New modes of similarity for complex data understanding
Applications of similarity-based operations
Image, video, voice, and music (multimedia) retrieval systems
Similarity for forensics and security

Important Dates

Abstract submission: May, 3rd, 2013
Paper submission: May, 10th, 2013
Notification: June, 21st, 2013
Final version: July, 10th, 2013
Conference: October 2-4, 2013

Submission guidelines

Papers submitted to SISAP 2013 must be written in English and formatted according to the LNCS guidelines. Full papers can be up to 12 pages, while short papers, case-studies/applications, and demos can be up to 6 pages (read below for types of contribution). By submitting a paper, its authors commit to have the paper presented at the conference by at least one of them if the paper is accepted.

Contributions

Authors are invited to submit previously unpublished papers on their research in the area of similarity search and applications. Papers should present original research contributions which bring out the importance of algorithms to applications. SISAP submissions can be of three kinds:

Research papers (full and short): SISAP accepts both full (12 pages) and short papers (6 pages). The full papers are expected to be descriptions of complete technical work, while the short papers will describe interesting, innovative ideas, which nevertheless require more work to mature – vision papers should also be submitted as short papers. All papers, regardless of size, will be given an entry in the conference proceedings.

Case-studies and applications: Submissions describe applications of existing similarity search technologies to interesting problems, including a description of the encountered challenges, how they were overcome, and the lessons learned. All papers on this track will be given an entry in the conference proceedings and a presentation slot, though the presentation slot duration may be shorter than for full research papers.

Demonstration papers: Submissions should provide the motivation for the demonstrated concepts, the information about the technology and the system to be demonstrated (including a system description, functionality and figures when applicable), and should state the significance of the contribution. Evaluation criteria for the demonstration proposals include: the novelty, the technical advances and challenges, and the overall practical attractiveness of the demonstrated system. Demonstration papers will also be given an entry in the conference proceedings – online demos are expected at the conference.

Program comittee chairs

Pavel Zezula, Masaryk University, Czech Republic
Nieves R. Brisaboa, Universidade da Coruna, Spain

Program comittee members

Giuseppe Amato, ISTI – Istituto di Scienza e Tecnologia dell’Informazione, Italy
Laurent Amsaleg, IRISA – Institut de Recherche en Informatique et Systemes Aleatoires, France
Benjamin Bustos, Universidad de Chile, Chile
Edgar Chavez, Universidad Michoacana, Mexico
Paolo Ciaccia, University of Bologna, Italy
Richard Connor, Strathclyde University, UK
Andrea Esuli, Istituto di Scienza e Tecnologie dell’Informazione, Italy
Rosalba Giugno, University of Catania, Italy
Michael Houle, National Institute of Informatics, Japan
Alexis Joly, INRIA, France
Daniel Keim, Universitat Konstanz, Germany
Eamonn Keogh, University of California – Riverside, USA
Magnus Lie Hetland, Norwegian University of Science and Technology, Norway
Yannis Manolopoulos, Aristotle University of Thessaloniki, Greece
Rui Mao, Shenzhen University, China
Luisa Mico, University of Alicante, Spain
Henning Muller, University of Applied Sciences Western Switzerland, Switzerland
Gonzalo Navarro, Universidad de Chile, Chile
Arlindo Oliveira, Lisbon Technical University, Portugal
Jose Oncina, University of Alicante, Spain
Apostolos Papadopoulos, Aristotle University of Thessaloniki, Greece
Marco Patella, University of Bologna, Italy
Vladimir Pestov, University of Ottawa, Canada
Matthias Renz, Ludwig-Maximilians-Universitat Munchen, Germany
Hanan Samet, University of Maryland, USA
Tomas Skopal, Charles University in Prague, Czech Republic
Bjorn Thor Jonsson, Reykjavik University, Iceland

Local organization

Nieves R. Brisaboa, Universidade da Coruna, Spain
Oscar Pedreira, Universidade da Coruna, Spain

March 6, 2013

PASCAL2

PASCAL2 IASD challenge: Deadline extension

=================================================================
Interactive Annotation of Sequential Data (IASD)
PASCAL2 challenge
http://translectures.eu/iasd
=================================================================
— Please, accept our apologies in case of multiple receptions —

*** New First Phase Deadline: March 25 ***

Dear colleagues,

We are pleased to announce an extension of the Interactive Annotation
of Sequential Data (IASD) PASCAL2 challenge deadlines. The aim of the
IASD challenge is to explore innovative, cost-effective solutions for
generating accurate transcriptions for video lectures from
VideoLectures.NET, and, at a more general level, speech-like
sequential data. The focus is not on developing advanced speech
recognition techniques, so much as on the study of techniques for
interacting intelligently with users. These techniques should seek to
optimise the trade-off between user effort and accuracy, in such a way
that the winning approaches are those reaching the top-3 accuracy with
minimum feedback.

Important Dates (extended):

Feb 12, 2013 – First challenge phase starts.
Mar 25, 2013 – First phase ends: selection of the 3 best systems.
Mar 26, 2013 – Second phase starts (for the 3 best systems only).
Apr 5, 2013 – Second phase ends: winners annouced and ranked.
Apr 6, 2013 – Reports deadline for the winners.

Presentation and awards:

The winners will be invited to present their systems at the joint
EUCOGIII/PASCAL meeting in Palma de Mallorca on April 11, 2013, with
travel costs covered by the meeting organisation. Winners attending
the meeting will be awarded with the following net prizes (after
taxes):

1st prize: €1000
2nd prize: €600
3rd prize: €300

You will find a detailed description of the challenge, data and
evaluation methodology at:

http://translectures.eu/iasd

Challenge Organizers:

Nicolas Serrano, Jesus Andres, Alfons Juan (UPV)
John Shawe-Taylor, Davor Orlic (K4A)
Mitja Jermol (JSI)

Challenge sponsors:

PASCAL2 Network (http://www.pascal-network.org)
EUCOGIII Network (http://www.eucognition.org)
transLectures project (http://translectures.eu)
Universitat Politecnica de Valencia (http://www.upv.es/index-en.html)
Knowledge for all (http://www.k4all.org)
Jožef Stefan Institute (http://www.ijs.si/ijsw/JSI)

March 5, 2013

PASCAL2

PhD studentship in computer vision, eye-tracking, and natural language, University of Edinburgh

University of Edinburgh
School of Informatics

PHD STUDENTSHIP IN COMPUTER VISION, EYE-TRACKING, AND NATURAL LANGUAGE

Applications are invited for a fully funded, three-year PhD studentship that combines ideas from computer vision, eye-tracking research, and natural language processing. The aim of the PhD project is to develop techniques for using human fixation data as recorded by an eye-tracker to train computer vision models, thus reducing the need for manual annotation. This approach will be augmented to exploit textual data (e.g., image captions) to improve object labeling. Another project aim is to crowd-source the eye-tracking data, e.g., through the use of webcams.

Applicants for the studentship must have:

* Strong undergraduate degree in computer science or a related
discipline

* Excellent programming skills

* Solid mathematical foundations (especially linear algebra and
probability theory)

* Fluency in English, both written and spoken

* UK or EU nationality — this is mandatory; applicants of other
nationalities will not be considered

* Master’s degree in a relevant area is desirable

* Experience in computer vision, machine learning, natural
language processing, or eye-tracking is desirable

The School of Informatics at Edinburgh is one of the top-ranked computer science departments in Europe and offers an exciting, interdisciplinary research environment. Edinburgh is a beautiful historic city with a high quality of life.

Starting date: September 2013 as soon as possible after that.

The PhD work will be carried out under the supervision of Dr. Vittorio Ferrari and Dr. Frank Keller, whose research interests can be found at:

http://groups.inf.ed.ac.uk/calvin/
http://homepages.inf.ed.ac.uk/keller/

For pre-screening, please send applications to the email address below, including:

* Full CV

* Full transcripts of both undergraduate and Master’s degree (if
applicable); this studentship requires high grades, especially
in mathematics and programming courses

* The names and email addresses of two referees

* List of publications, if you have prior research experience

Contact: Vittorio Ferrari, vferrari@staffmail.ed.ac.uk

Deadline: 25 April 2013

March 5, 2013

PASCAL2

BioASQ challenge on large-scale biomedical semantic indexing and question answering

BioASQ challenge on large-scale biomedical semantic indexing and question
answering
(BioASQ workshop to be collocated with CLEF 2013 in Valencia, Spain on
September 27, 2013)

Web site: http://bioasq.org/
twitter: https://twitter.com/bioasq

The BioASQ challenge aims to push for a solution to the information access
problem of biomedical experts. It will evaluate the ability of systems to
perform various tasks in the biomedical QA process:
1. large-scale classification of biomedical documents onto ontology concepts
(semantic indexing),
2. classification of biomedical questions onto relevant concepts,
3. retrieval of relevant document snippets, concepts and knowledge base
triples, and
4. delivery of the retrieved information in a concise and
user-understandable form.

In particular, the challenge will comprise two main tasks:

Task 1a. Large-scale online biomedical semantic indexing
Large-scale semantic indexing will be evaluated on the whole of PubMed. In
particular, participants will be asked to classify incoming documents before
the human curators do:
* BioASQ will distribute new unclassified PubMed documents.
* Participants will have a limited response time to attach MeSH terms.

Task 1a will run for three consecutive periods (batches) of 6 weeks each.
The first batch will start on April 15, 2013. Separate winners will be
announced for each batch. Participation in the task can be partial, i.e. in
some of the batches.

Task 1b. Biomedical semantic QA
The systems will be evaluated against gold answers created by a team of
biomedical experts. The task will run in two phases:
In Phase A:
* BioASQ will transmit simultaneously questions from the benchmark.
* Participants will have limited time to respond with concepts, snippets,
triples.
In Phase B:
* BioASQ will distribute questions + concepts, snippets, triples.
* Participants will respond with facts, summaries, etc.

The two phases will run consecutively, i.e. Phase B will start after the end
of phase A. Phase A will start on June 3, 2013. Separate winners will be
announced for each phase and each task target, i.e. concepts, snippets,
triples, etc. Participation in the task can be partial, i.e. in one of the
two phases and only for some of the targets.

Prizes will be awarded to the winners. Details about the prizes will be
announced on the Web site of the challenge.

Important dates:
March 18, 2013: Training data available for task 1a. Dry-run data available
for task 1b.
April 15, 2013: Start of task 1a.
June 3, 2013: Start of task 1b.
July 15, 2013: Submission of papers for BioASQ workshop
August 30, 2013: End of challenge
September 27, 2013: BioASQ workshop, collocated with CLEF 2013

The BioASQ challenge and workshop are organised by the BioASQ project,
supported by the European Commission within the 7th Framework Programme
(Grant Agreement No. 318652).

March 5, 2013

PASCAL2

CFP: ICML 2013 Workshop on Spectral Learning

Call for Papers: Workshop on Spectral Learning — @ICML2013
June 20 or 21, Atlanta (GA), USA
Website: http://sites.google.com/site/spectrallearningworkshop/

Many problems in machine learning involve collecting high-dimensional multivariate observations or sequences of observations, and then fitting a compact model which explains these observations. Recently, linear algebra techniques have given a fundamentally different perspective on how to fit and perform inference in these models. Exploiting the underlying spectral properties of the model parameters has led to fast, provably consistent methods for parameter learning that stand in contrast to previous approaches, such as Expectation Maximization, which suffer from slow convergence and issues related to local optima.

In the past several years, these spectral learning algorithms have become increasingly popular. They have been applied to learn the structure and parameters of many models including predictive state representations, finite state transducers, hidden Markov models, latent trees, latent junction trees, probabilistic context free grammars, and mixture/admixture models. Spectral learning algorithms have also been applied to a wide range of application domains including system identification, video modeling, speech modeling, robotics, and natural language processing.

The focus of this workshop will be on spectral learning algorithms, broadly construed as any method that fits a model by way of a spectral decomposition of moments of (features of) observations. We would like the workshop to be as inclusive as possible and encourage paper submissions and participation from a wide range of research related to this focus. This includes (but is not limited to):

* Linear-algebraic methods for estimation and inference in probabilistic models and weighted automata/operator models
* Spectral approaches to dimension reduction (e.g., with applications in estimating mixture models)
* Method-of-moment estimation via higher-order tensor decompositions
* Spectral graph theory and applications in clustering and learning on manifolds
* Domain-specific aspects of using spectral approaches in applications

Submitted papers should be in the ICML 2013 format with a maximum of 4 pages (not including references). Please e-mail your submission to spectralicml2013@gmail.com with the subject line “Submission to Spectral Learning Workshop”. Contributions will be considered for both short talks and poster presentations.

Concurrent submissions to the workshop and the main conference (or other conferences) are permitted.

Important dates:

* Submission deadline: April 6, 2013
* Notification of acceptance: April 20, 2013 (tentative)
* Workshop: June 20 or 21, 2013

Organizers:

* Byron Boots (University of Washington)
* Daniel Hsu (Microsoft Research New England)
* Borja Balle (Universitat Politècnica de Catalunya)
* Ankur Parikh (Carnegie Mellon University)

November 30, 2011

PASCAL2

Funded PhD studentships, School of Informatics, Edinburgh

We would like to advertise the following studentships starting in 2012. Please circulate to those who might be interested.

Microsoft Research Funded PhD Scholarship in Machine Learning Markets
Supervisor: Amos Storkey
a.storkey(at)ed.ac.uk

Microsoft Research Funded PhD Scholarship in Statistical Machine Learning and Natural Language Processing of Programming Language Text
Supervisor: Charles Sutton
csutton(at)inf.ed.ac.uk

School of Informatics, University of Edinburgh
Duration of Studentships: 36 months
Preliminary deadline: 16 December 2011

For further information see

http://homepages.inf.ed.ac.uk/amos/jobsphds.html

http://homepages.inf.ed.ac.uk/csutton/projects.html#nlppl

November 24, 2011

PASCAL2

Announcing the PASCAL Heart Sounds Challenge

We are pleased to announce the PASCAL-sponsored Heart Sounds Challenge. Here is your chance to prove your machine learning technique can outperform those of everyone else – and win an iPad for your efforts! (Also come to the Canary Islands to present your results in a workshop after AISTATS!)

For more details see: http://www.peterjbentley.com/heartchallenge/

According to the World Health Organisation, cardiovascular diseases (CVDs) are the number one cause of death globally: more people die annually from CVDs than from any other cause. An estimated 17.1 million people died from CVDs in 2004, representing 29% of all global deaths. Of these deaths, an estimated 7.2 million were due to coronary heart disease. Any method which can help to detect signs of heart disease could therefore have a significant impact on world health. This challenge is to produce methods to do exactly that. Specifically, we are interested in creating the first level of screening of cardiac pathologies both in a Hospital environment by a doctor (using a digital stethoscope) and at home by the patient (using a mobile device).

For this challenge we have two datasets comprising several hundred real heart sounds, gathered from an iphone app by the general public, and by a digital stethoscope in a noisy hospital environment.

Challenge 1 is segmentation – can your method correctly identify the “lub dub” (S1 and S2) components of the sound?

Challenge 2 is classification – can your method correctly classify the heart sounds into categories such as Normal, Murmur, Extra Heart Sound, and Artifact?

This problem is of particular interest to machine learning researchers as it involves classification of audio sample data, where distinguishing between classes of interest is non-trivial. Data is gathered in real-world situations and frequently contains background noise of every conceivable type. The differences between heart sounds corresponding to different heart symptoms can also be extremely subtle and challenging to separate. Success in classifying this form of data requires extremely robust classifiers. Despite its medical significance, to date this is a relatively unexplored application for machine learning.

Enquiries and submission, email: Yiqi Deng, y.deng.11(at)ucl.ac.uk

November 22, 2011

PASCAL2

CFP: British Machine Vision Conference 2012

BMVC 2012: British Machine Vision Conference, University of Surrey, UK Sept 3-7th 2012

CALL FOR PARTICIPATION

http://bmvc2012.surrey.ac.uk/

The British Machine Vision Conference (BMVC) is one of the major international conferences on machine vision and related areas. Organized by the British Machine Vision Association, the 23rd BMVC will be held in Guildford UK, at the University of Surrey.

Authors are invited to submit full-length high-quality papers in image processing and machine vision. Papers covering theory and/or application areas of computer vision are invited for submission. Submitted papers will be refereed on their originality, presentation, empirical results, and quality of evaluation.

All papers will be reviewed *doubly blind*, normally by three members of our international programme committee. Please note that BMVC is a single track meeting with oral and poster presentations and will include two keynote presentations and two tutorials.

Topics include, but are not limited to:

• Statistics and machine learning for vision
• Stereo, calibration, geometric modelling and processing
• Person, face and gesture tracking
• Object and activity recognition
• Motion, flow and tracking
• Segmentation and feature extraction
• Model-based vision
• Image processing techniques and methods
• Texture, shape and colour
• Video analysis
• Document processing and recognition
• Vision for quality assurance, medical diagnosis, etc.
• Vision for visualization, interaction, and graphics

Conference Chairs:
Dr John Collomosse
Dr Krystian Mikolajczyk
Prof Richard Bowden

Important Dates
26 April 2012 Abstracts due
3 May 2012 Full paper submissions due
14 June 2012 Deadline for return of reviews
2 July 2012 Area chair recommendations due
6 July 2012 Author notifications
1 August 2012 Camera ready papers due
3-7 September 2012 Conference

See http://bmvc2012.surrey.ac.uk/ for more details