i++ Departmental Newsletter
Week commencing 22 September 2008
Previous Newsletters
IEEE Transactions on Computational Intelligence and AI in Games - New
Journal Launched
The IEEE Transactions on Computational Intelligence and AI in Games opened
for submissions earlier this week, ready for the launch issue in March 2009.
Games provide a superb test-bed and application for all kinds of Computational
Intelligence and AI research, and having an IEEE Transactions to provide a focus
for high quality work in the area is a major step forward.
Dr. Simon Lucas, the founding editor-in-chief commented: “The idea of the
journal was first discussed,
at the IEEE Symposium Series on Computational Intelligence in Honolulu, April
2007. It is fantastic that less than 18 months later we now have the journal
open and ready for business. The next major milestone is getting the
inaugural issue out on time, and full of high-quality high-impact papers, but it
is already looking very promising.
The journal has a broad focus, covering video games, traditional board games,
and applications of games (and the associated AI) to real world problems, such
as trying to avoid financial meltdowns.
The pace of progress in this area is astonishing. For example, until very
recently many experts in the area argued that computers would never play a
strong game of Go, yet now we have Monte-Carlo players such as MoGo worrying the
leading professional players. These algorithms also have applications well
beyond board games, and the whole area is ripe for more interdisciplinary
research. A high quality journal has an important role to play in fostering
this.”
Natural Language Engineering Group Language and Computation Day
The Natural Language Engineering (NLE) Group in the department continues to
grow with another Senior Research Officer and two more PhD students expected to
join in October. The annual Language and Computation Day is being held on Friday
3 October 2008 so you can learn more about the group's activities and ongoing
research. This day is organised by the
Language and Computation Group, an interdisciplinary group created to foster
interaction between researchers within the University. It includes staff and
students from the NLE group as well as the Data Archive and the Department of
Languages and Linguistics.
The programme has not been finalised because the organisers are keen to present
the latest research and results; however confirmed talks include presentations
by Doug Arnold, Sonja Eisenbeiss, Udo Kruschwitz, Deirdre Lungley, Jon
Chamberlain and Richard Sutcliffe.
For more information about the LAC day please visit the
Language and Computation Day website.
AnaWiki Project Launches Phrase Detectives Game
The
AnaWiki project, based at Essex University, has released the first stage of its
website to collect linguistic information from Web users. The website, called
Phrase Detectives, is a fun game where users make annotations in documents
including short stories, news, travel advice and text written by other users.
The most points are scored when there is agreement between users. The aim of the
project is to collect a significant amount of data and investigate the
possibility of using mass collaboration to train computer systems.
To play the game and find out more information visit the
Phrase Detectives website.
Summer Keynote Speeches
Professor Richard Bartle has given several keynote speeches over the summer
period: Virtual Policy 08, BERR (Dept. for Business Enterprise and Regulatory
Reform - formerly DTI); Digital Interactive Symposium, Edinburgh Interactive
Festival and Postgraduate Games Conference 08, Brunel University.
Also, although not a keynote, Richard's June
interview with the Massively online games news website was voted their most
important of the year.
Papers Published
Y. Sun, A. Erol, M Yilmaz, M. C. Arikan,B. Ulug, A. Ulug, N. Balkan, M.
Sopanen, O.Reentilä, M. Mattila, C. Fontaine, and A. Arnoult, Optical
and electrical properties of modulation-doped n and p type GaxIn1-xNyAs1-y/GaAs
quantum wells for 1.3 um laser applications, doi:
10.1007/s11082-007-9163-8, Optical and Quantum Electronics, (2007)
Y. Sun, A. Erol, N. Balkan, M. C. Arikan, Electronic transport in n- and
p-type modulation doped GaInNAs/GaAs quantum wells,
doi:10.1016/j.mejo.2008.06.010, Microelectronics Journal (2008)
S. B. Lisesivdin, N. Balkan, E. Ozbay, A Simple Parallel Conduction
Extraction Method (SPCEM) for MODFETs and Undoped GaN-based HEMTs,
doi:10.1016/j.mejo.2008.06.006, Microelectronics Journal (2008)
B. Royall, N. Balkan, Dilute Nitride N-I-P-I Solar Cells,
doi:10.1016/j.mejo.2008.06.011, Microelectronics Journal (2008)
M. Yilmaz, Y. Sun, N. Balkan, B. Ulug, A. Ulug, M Sopanen, O. Reentilä, M.
Mattila, C. Fontaine, and A. Arnoult, Photoluminescence in n and p
Modulation Doped GaInNAs/GaAs Quantum Wells,
doi:10.1016/j.mejo.2008.06.003, Microelectronics Journal (2008)
S. Chung, N. Balkan, The Anatomy of the Gunn Laser, Journal of
Applied Physics Volume 104 Issue 6 (2008)
N. Balkan, S. H. Chung, Applications of Gunn Lasers, Photonics
Europe (2008), Proceedings of SPIE Volume 6997, 699706-1 (2008)
_______________________________________________________________________________________
M.A. Oskoei and H. Hu, Support Vector Machine Based Classification
Scheme for Myoelectric Control Applied to Upper Limb, IEEE Transactions
on Biomedical Engineering, Vol. 55, No. 8, 2008, pages 1956-1965.
Abstract - This paper proposes and evaluates the application of support
vector machine (SVM) to classify upper limb motions using myoelectric signals.
It explores the optimum configuration of SVM-based myoelectric control, by
suggesting an advantageous data segmentation technique, feature set, model
selection approach for SVM, and post-processing methods. This work presents a
method to adjust SVM parameters before classification, and examines overlapped
segmentation and majority voting as two techniques to improve controller
performance. A SVM, as the core of classification in myoelectric control, is
compared with two commonly used classifiers: linear discriminant analysis (LDA)
and multilayer perceptron (MLP) neural networks. It demonstrates exceptional
accuracy, robust performance, and low computational load. The entropy of the
output of the classifier is also examined as an online index to evaluate the
correctness of classification; this can be used by online training for long-term
myoelectric control operations.
R. Samperio, H. Hu, F. Martin and V. Matellan, A hybrid approach to fast
and accurate localisation for legged robots, International Journal of
Robotica, Vol. 26, 2008, pages 1-14.
Abstract - This paper describes a hybrid approach to a fast and accurate
localisation method for legged robots based on Fuzzy-Markov (FM) and Extended
Kalman Filters (EKF). Both FM and EKF techniques have
been used in robot localisation and exhibit different characteristics in terms
of processing time, convergence and accuracy. We propose a Fuzzy-Markov-Kalman
(FM-EKF) localisation method as a combined solution for a poor predictable
platform such as Sony Aibo walking robots. The experimental results show the
performance of EKF, FM and FM-EKF in a localisation task with simple movements,
combined behaviours and kidnapped situations. An overhead tracking system was
adopted to provide a ground truth to verify the performance of the proposed
method.
Y. Tao and H. Hu, A novel sensing and data fusion system for 3D arm
motion tracking in tele-rehabilitation, IEEE Transactions on
Instrumentation and Measurement, Vol. 57, No. 5, pages 1029-1040, 2008
Abstract - In this paper, we present a novel sensing and data fusion system
to track 3D arm motion in a tele-rehabilitation program. A Particle Filter (PF)
algorithm is adopted in the system to fuse data from inertial and visual sensors
in a probabilistic manner. It is able to propagate multi-modal distributions of
system states based on an “importance sampling” technique, using sets of
weighted particles. To avoid the problem of conventional PF algorithms that
suffer from particle degeneracy and perform poorly in a narrow distribution
situation, we adopt two strategies in our system; namely state space pruning and
an arm physical geometry constraint. Experimental results show that the proposed
PF framework outperforms other fusion methods and provides accurate results in
comparison to the ground truth.
Forthcoming Seminar
Dr Selmer Bringsjord, Rensselaer Polytechnic Institute (USA) (hosted by Dr
Amnon Eden)
10 October, 3.00-4.00pm, Room 1N1.4.1
Doing AI That’s Tough Enough: Human Genius, Hypercomputation, and Automatic
Programming.
Abstract - Human persons are astonishingly smart. Members of this group
have managed to produce such things as Hamlet, G\"{o}delian
incompleteness results, and the Ninth Symphony; we could continue the list
indefinitely. On the other hand, many of the things humans can do don’t
require lots of smarts. (Ants can walk from point A to B, and when humans
do such things, in general, they do nothing ingenious.) Unfortunately, AI
has devolved away from striving to engineer computing machines that are
human-level smart. Dr Bringsjord explains this situation and makes it
precise by explicating ‘human-level smart’ as doing something that requires
hypercomputation. He will present the latest example (within the context of
prior ones) of AI work he is pursuing because, Dr Bringsjord writes, "it’s
truly tough enough: automatic programming."
