Tuesday, March 30, 2010

Playing “Pong” with the blink of an eye: low cost technology could enable people with severe physical disabilities to become ‘gamers’ for the first time

Imperial College London News Release

University students have developed a computer game that is operated by eye movements, which could allow people with severe physical disabilities to become ‘gamers’ for the first time, they announce today.

The students, from Imperial College London, have adapted an open source game called ‘Pong’, where a player moves a bat to hit a ball as it bounces around the screen. The adaptation enables the player to move the bat using their eye.

To play the game, the user wears special glasses containing an infrared light and a webcam that records the movement of one eye. The webcam is linked to a laptop where a computer program syncs the player’s eye movements to the game.

The prototype game is very simple but the students believe that the technology behind it could be adapted to create more sophisticated games and applications such as wheelchairs and computer cursors controlled by eye movements.

One of the major benefits of the new technology is that it is inexpensive, using off-the-shelf hardware and costing approximately £25 to make. Eye movement systems that scientists currently use to study the brain and eye motion cost around £27,000, say the researchers.

Dr Aldo Faisal, the team’s supervisor from the Department of Computing and the Department of Bioengineering at Imperial College London, says:

“Remarkably, our undergraduates have created this piece of neurotechnology using bits of kit that you can buy in a shop, such as webcams. The game that they’ve developed is quite simple, but we think it has enormous potential, particularly because it doesn’t need lots of expensive equipment. We hope to eventually make the technology available online so anyone can have a go at creating new applications and games with it and we’re optimistic about where this might lead. We hope it could ultimately provide entertainment options for people who have very little movement. In the future, people might be able to blink to turn pages in an electronic book, or switch on their favourite song, with the roll of an eye.”

Mr Ian Beer, who is a third year undergraduate from the Department of Computing, adds: “This game is just an early prototype, but we’re really excited that from our student project we’ve managed to come up with something that could ultimately help people who have really limited movement. It would be fantastic to see lots of people across the world creating new games and applications using our software.”

Researchers in Dr Faisal’s lab are now refining the technology so that it can monitor movements in both eyes. This would enable a user to carry out more complicated tasks such as plotting a journey on screen. This might ultimately allow them to use eye movements to steer a motorised wheelchair.

For further information please contact:

Colin Smith
Press Officer
Imperial College London
Email: cd.smith@imperial.ac.uk
Tel: +44 (0)207 594 6712
Out of hours duty press officer: +44 (0)7803 886 248

Notes to Editor

1. Student team includes: William Abbot, Department of Bioengineering; Oliver Rogers, Department of Maths and Department of Computing; Tim Treglown, Department of Maths and Department of Computing; Aaron Berk, Department of Computing; Ian Beer, Department of Computing.

2. About Imperial College London

Consistently rated amongst the world's best universities, Imperial College London is a science-based institution with a reputation for excellence in teaching and research that attracts 14,000 students and 6,000 staff of the highest international quality.

Innovative research at the College explores the interface between science, medicine, engineering and business, delivering practical solutions that improve quality of life and the environment - underpinned by a dynamic enterprise culture.

Since its foundation in 1907, Imperial's contributions to society have included the discovery of penicillin, the development of holography and the foundations of fibre optics. This commitment to the application of research for the benefit of all continues today, with current focuses including interdisciplinary collaborations to improve health in the UK and globally, tackle climate change and develop clean and sustainable sources of energy.

Website: www.imperial.ac.uk

Saturday, March 27, 2010

Topic models for image retrieval on large-scale databases

Eva Hörster, Rainer Lienhart, Wolfgang Effelsberg, Bernhard Möller

With the explosion of the number of images in personal and on-line collections, efficient techniques for navigating, indexing, labeling and searching images become more and more important. In this work we will rely on the image content as the main source of information to retrieve images. We study the representation of images by topic models in its various aspects and extend the current models. Starting from a bag-of-visual-words image description based on local image features, images representations are learned in an unsupervised fashion and each image is modeled as a mixture of topics/object parts depicted in the image. Thus topic models allow us to automatically extract high-level image content descriptions which in turn can be used to find similar images. Further, the typically lowdimensional topic-model-based representation enables efficient and fast search, especially in very large databases.


Friday, March 26, 2010

14th Pan-Hellenic Conference on Informatics with international Participation

2nd Call for Papers and Deadline Extension

September 10-12, 2010, Tripoli, Greece

The Greek Computer Society (ΕΠΥ), the Department of Computer Science and Technology, University of Peloponnese, and the Department of Telecommunications Science and Technology, University of Peloponnese organize the 14th Panhellenic Conference on Informatics (PCI 2010) at Tripoli, Greece, during 10 - 12 of September, 2010.

The PCI is an event established by the Greek Computer Society. The 1st Conference took place at Athens (1984), the 2nd at Thessaloniki (1988), the 3rd at Athens (1991), the 4th at Patras (1993), the 5th at Athens (1995), the 6th at Athens (1997), the 7th at Ioannina (1999), the 8th at Nicosia Cyprus (2001), the 9th at Thessaloniki (2003), the 10th at Volos (2005), the 11th at Patras (2007), the 12th at Samos (2008) and the 13th at Corfu Island (2009).

Topics of Interest

The PCI 2010 Conference will run in parallel sessions, with invited talks, research and case study tracks. Authors are invited to submit papers in any area of Informatics, Computer Science, Computer Engineering, Telecommunications, and Information Systems. Topics of interest include, but are not limited to, the following:

· Algorithms and Data Structures

· Artificial Intelligence

· Bioinformatics

· Business Intelligence

· Communication and Information Systems Security

· Computational Science

· Computer and Communication Networks

· CRM and ERP systems

· Cultural and Museum Information Systems

· Databases

· Data Mining

· Digital Libraries

· eCommerce, eBusiness, eGovernment, eHealth

· Education Technologies

· Graphics, Visualization, Multimedia and Virtual Reality

· Grid and Cluster Computing

· Hardware and Architecture

· Human-Computer Interaction

· Image and Video Processing

· Information Retrieval

· Information Society: Legal and Regulatory issues

· Information Systems

· Information Theory

· Knowledge based Systems

· Open Source Software

· Operating Systems

· Parallel and Distributed Systems

· Privacy and Privacy Enhancing Technologies

· Programming Languages

· Sensor Networks

· Signal Processing

· Software Engineering

· Telecommunication Systems and Policies

· Wearable Computing

· Wireless and Mobile Computing

Important dates

Full paper submission deadline:

16 April 2010

Notification of decision:

11 May 2010

Camera-ready deadline:

11 June 2010

Authors’ registration deadline:

11 June 2010

WIE 2010 submission deadline:

30 April 2010

Instructions for Authors

Authors are invited to submit original manuscripts, in English, limited in length to five (5) pages. The required format is IEEE double‐column (available in doc and LaTeX format at http://www.ieee.org/portal/pages/pubs/transactions/stylesheets.html). All submitted papers will undergo a peer review process, coordinated by the PC Chairs. The reviewers are not required to read any appendices; the paper should be intelligible on its own. Authors are invited to send their manuscripts electronically in Postscript or PDF format at the PCI Conference website http://pci2010.uop.gr

Submission of a paper implies the willingness of at least one of the authors to register and present the paper. Each participant may be associated with at most two papers.

PCI 2010 Proceedings

The PCI Conference proceedings will be published by IEEE Computer Society, Conference Publishing Services and distributed at the conference. IEEE Conference Publishing Services arranges for indexing through Thomson ISI, IEE (INSPEC), EI (Compendex), and other indexing services and archives the publication to IEEE‐Xplore and the IEEE Computer Society Digital Libraries (CSDL).


Language of the PCI 2010 Conference

The PCI Conference is mainly a forum for information exchange between Greek scientists from Greece, Cyprus and the Diaspora with international participation. So, English will be the official language of the PCI Conference, including presentations and proceedings. Non Greek‐speaking authors are strongly encouraged to submit papers and participate in the PCI 2010 Conference, as well.


Thursday, March 25, 2010

My Erdös number is 4!!!

The Erdős number describes the "collaborative distance" between a person and mathematician Paul Erdős, as measured by authorship of mathematical papers.

It was created by friends as a humorous tribute to the enormous output of Erdős, one of the most prolific modern writers of mathematical papers, and has become well-known in scientific circles as a tongue-in-cheek measurement of mathematical prominence.

Paul Erdős was an influential and itinerant mathematician, who spent a large portion of his later life living out of a suitcase and writing papers with those of his colleagues willing to give him room and board. He published more papers during his life than any other mathematician in history (at least 1400).

To be assigned an Erdős number, an author must co-write a mathematical paper with an author with a finite Erdős number. Paul Erdős is the one person having an Erdős number of zero. For any author other than Erdős, if the lowest Erdős number of all of his coauthors is k, then the author's Erdős number is k + 1.

Erdős wrote around 1,500 mathematical articles in his lifetime, mostly co-written. He had 511 direct collaborators; these are the people with Erdős number 1. The people who have collaborated with them (but not with Erdős himself) have an Erdős number of 2 (8,162 people as of 2007), those who have collaborated with people who have an Erdős number of 2 (but not with Erdős or anyone with an Erdős number of 1) have an Erdős number of 3, and so forth. A person with no such coauthorship chain connecting to Erdős has an Erdős number of infinity (or an undefined one).

There is room for ambiguity over what constitutes a link between two authors; the Erdős Number Project Web site says "Our criterion for inclusion of an edge between vertices u and v is some research collaboration between them resulting in a published work. Any number of additional co-authors is permitted," but they do not include non-research publications such as elementary textbooks, joint editorships, obituaries, and the like. The “Erdős number of the second kind” restricts assignment of Erdős numbers to papers with only two collaborators.

Article from wikipedia

My Erdös number is 4 via the link: Savvas A. Chatzichristofis (2010) -> Avi Arampatzis (2004) -> Marc J. van Kreveld (1998) -> Boris Aronov (1994) -> Paul Erdös !


Wednesday, March 24, 2010

Content-Aware Fill Sneak Peek

One of the biggest requests we get of Photoshop is to make adding, removing, moving or repairing items faster and more seamless. From retouching to completely reimagining an image, heres an early glimpse of what could happen in the future when you press the delete key. How might you use this new capability in your workflow?

Monday, March 22, 2010

Consumer Image Retrieval by Estimating Relation Tree From Family Photo Collections

Zhang, Tong; Chao, Hui; Willis, Chris; Tretter, Dan
HP Laboratories


Keyword(s): Consumer image retrieval, family relation tree, face analysis, face clustering, demographic assessment, and social networking

Abstract: In this paper, we propose an approach to automatically estimate relationship among people in a family image collection based on results from face analyses technologies including automated face recognition and clustering, demographic assessment, and face similarity measurement, as well as contextual information such as people co-appearance, people's relative positions in photos and image timestamps. As the result, a relation tree can be estimated which provides important semantic information regarding people involved in a photo collection and has numerous applications in photo sharing and browsing, social networking, etc. The methods for deriving and integrating information from photos and the process for estimating a relation tree are described. Experimental results on two typical consumer photo collections and examples of using these results in consumer image retrieval are presented.


The Key To Gmail: Sh*t Umbrellas

Today at the Gmail Behind The Scenes panel at the SXSW festival in Austin, Texas, key team members of the Gmail team revealed the true secret of the service: Shit umbrellas.

Product manager Todd Jackson made the humorous revelation when explaining how the Gmail team works as a group of about 100 people, the vast majority of which are engineers. “You can either be a shit funnel or a shit umbrella,” Jackson says.

What he means by that is that as a product with hundreds of millions of users (and a company with thousands of employees) there’s a lot of stuff constantly being hurled at the team — as a shit umbrella, the product managers protect the engineers from getting distracted. It’s not enough to be a “shit funnel” where they would pass some of the junk down to engineers, they need to fully protect the engineers.

This sentiment was echoed by Edward Ho, who is known as “Mr. Buzz,” as he’s the one who built up the Google Buzz team (a sub-unit of the Gmail team). Ho noted his hatred for unnecessary meetings, and has made sure that when the Buzz team needs to have them, they are based around demos, not talking about things. “It’s all about what you’ve done,” Ho says.

Some other interesting notes about Gmail:

  • The original invites system wasn’t a marketing ploy, it was simply an engineering decision to make sure they could scale
  • There’s a 30-1 engineers to products managers ratio in the Gmail team — it’s certainly one of the biggest ratios at Google
  • The Gmail team is spread over a few offices around the world (including Zurich), it used to be more, but they consolidated to help the product.
  • There are “hundreds of million of users” — the third-largest email provider
  • In India, Gmail is the number one email provider
  • Gmail is growing fasters internationally than in the U.S.
  • Gmail is available in 53 languages
  • Internally, the Google Buzz team was known as “Team Taco Town” after an SNL skit
  • Google uses Gmail internally (obviously), switched over from Microsoft Outlook at launch (about 6 years ago)
  • Gmail is slow for some users mainly because they have a ton of emails saved. A fix for that is coming soon
  • Most of gmail is written in Java, JavaScript, C++
  • There are several hundred thousands lines of javascript in Gmail – one of the biggest in the world
  • No new feature can launch for Gmail that adds latency to the product

[photo: flickr/atomicjeep]

Article From http://techcrunch.com/2010/03/14/key-to-gmail/

Saturday, March 20, 2010

Search through a million images in less than a second

Article From Mathias Lux

We reached a milestone here! Sebastian Kielmann reported that he used LIRe to index a million images. While that’s actually no problem he further managed to search through the images in < 1 second! Whoot!

Sebastian used the metric index, which implements the ideas of Giuseppe Amato. The approach is easy, but works out really fine. Currently CEDD is the standard descriptor, but others are integrated easily.  However, using the metric index is not trivial and requires some knowledge on the process. Also the results are approximate and might differ from the results obtained by linear search.


General Motors Augmented Reality Windshield

General Motors has developed a working next-generation heads-up display that turns an ordinary windshield into an augmented reality information dashboard. Such a system can improve safety and advance knowledge behind the wheel, visually identifying important objects in physical space like road signs, the edges of the road you're on in conditions of poor visibility like fog, and even bring GPS functions right into the dashboard by outlining the exact building you're going to.

Friday, March 19, 2010

SPIRS: A Web-based image retrieval system for large biomedical databases

William Hsu, Sameer Antani, L. Rodney Long, Leif Neve and George R. Thoma


With the increasing use of images in disease research, education, and clinical medicine, the need for methods that effectively archive, query, and retrieve these images by their content is underscored. This paper describes the implementation of a Web-based retrieval system called SPIRS (Spine Pathology & Image Retrieval System), which permits exploration of a large biomedical database of digitized spine X-ray images and data from a national health survey using a combination of visual and textual queries.


SPIRS is a generalizable framework that consists of four components: a client applet, a gateway, an indexing and retrieval system, and a database of images and associated text data. The prototype system is demonstrated using text and imaging data collected as part of the second U.S. National Health and Nutrition Examination Survey (NHANES II). Users search the image data by providing a sketch of the vertebral outline or selecting an example vertebral image and some relevant text parameters. Pertinent pathology on the image/sketch can be annotated and weighted to indicate importance.


During the course of development, we explored different algorithms to perform functions such as segmentation, indexing, and retrieval. Each algorithm was tested individually and then implemented as part of SPIRS. To evaluate the overall system, we first tested the system's ability to return similar vertebral shapes from the database given a query shape. Initial evaluations using visual queries only (no text) have shown that the system achieves up to 68% accuracy in finding images in the database that exhibit similar abnormality type and severity. Relevance feedback mechanisms have been shown to increase accuracy by an additional 22% after three iterations. While we primarily demonstrate this system in the context of retrieving vertebral shape, our framework has also been adapted to search a collection of 100,000 uterine cervix images to study the progression of cervical cancer.


SPIRS is automated, easily accessible, and integratable with other complementary information retrieval systems. The system supports the ability for users to intuitively query large amounts of imaging data by providing visual examples and text keywords and has beneficial implications in the areas of research, education, and patient care

Read More

Lire 0.8 released

The LIRE (Lucene Image REtrieval) library provides a simple way to create a Lucene index of image features for content-based image retrieval (CBIR), which allows searching for similar images.

A major change in this version is the support of Lucene 3.0.1, which has a changed API and better performance on some operating systems. A critical bug was fixed in the Tamura feature implementation. It now definitely performs better. Hidden in the depths of the code, there is an implementation of the approximate fast indexing approach of G. Amato. It copes with the problem of linear search and provides a method for fast approximate retrieval for huge repositories (millions?).

Download Lire & Lire Demo

Tuesday, March 9, 2010

MammoSys: A content-based image retrieval system using breast density patterns

Júlia E.E. de Oliveira, Alexei M.C. Machado, Guillermo C. Chavez, Ana Paula B. Lopes, Thomas M. Deserno, Arnaldo de A. Araújo


In this paper, we present a content-based image retrieval system designed to retrieve mammographies from large medical image database. The system is developed based on breast density, according to the four categories defined by the American College of Radiology, and is integrated to the database of the Image Retrieval in Medical Applications (IRMA) project, that provides images with classification ground truth. Two-dimensional principal component analysis is used in breast density texture characterization, in order to effectively represent texture and allow for dimensionality reduction. A support vector machine is used to perform the retrieval process. Average precision rates are in the range from 83% to 97% considering a data set of 5024 images. The results indicate the potential of the system as the first stage of a computer-aided diagnosis framework.

Keywords: Medical images; Breast density; Content-based image retrieval; Two-dimensional principal component analysis; Support vector machine

Read More

Satellite Image Retrieval Based On Ontology Merging

Friday, March 5, 2010

Image retrieval from the web using multiple features

Author(s): A. Vadivel, Shamik Sural, A.K. Majumdar
Journal: Online Information Review

Purpose – The main obstacle in realising semantic-based image retrieval from the web is that it is difficult to capture semantic description of an image in low-level features. Text-based keywords can be generated from web documents to capture semantic information for narrowing down the search space. The combination of keywords and various low-level features effectively increases the retrieval precision. The purpose of this paper is to propose a dynamic approach for integrating keywords and low-level features to take advantage of their complementary strengths.
Design/methodology/approach – Image semantics are described using both low-level features and keywords. The keywords are constructed from the text located in the vicinity of images embedded in HTML documents. Various low-level features such as colour histograms, texture and composite colour-texture features are extracted for supplementing keywords.
Findings – The retrieval performance is better than that of various recently proposed techniques. The experimental results show that the integrated approach has better retrieval performance than both the text-based and the content-based techniques.
Research limitations/implications – The features of images used for capturing the semantics may not always describe the content.
Practical implications – The indexing mechanism for dynamically growing features is challenging while practically implementing the system.