Updates from Klagenfurt

It's been a while since my last post... Time for some updates...

My stay at Klagenfurt University is quickly approaching its end. It has been a very successful period, by every measure. Here are some highlights since my last posting:

• Last Friday we had the last day of classes, including presentations and project demonstrations by students. They did a superb job using a variety of languages, tools, and techniques to create Visual Information Retrieval (VIR) solutions and produced excellent prototypes and reports. 
  
• Four PhD students in the class expressed interest in continuing working with me on focused topics that build on the knowledge they acquired during the course. This is excellent news and a good sign of high-quality demos and papers within a few months.
• Mathias Lux and I are busy preparing our second paper together in which we propose an innovative approach for tag classification and noisy tag removal in social image sharing sites such as flickr.
• I am also very busy putting the final touches on my technical talk on "Image retrieval using computational models of visual attention" this Friday. It is part of a mini-workshop that starts with a talk by Ansgar Scherp (whom I first met in Innsbruck back in March) titled "Towards an Ecosystem for Semantics".
• All around Klagenfurt, the expectations for the beginning of the UEFA EuroCup 2008 are running high. The competition starts on Saturday and the first game in Klagenfurt will happen on Sunday night.

First joint paper!

My colleague Mathias Lux, his student Arthur Pitman, and I have just finished (in record time!) our first joint paper: "Using Visual Features to Improve Tag Suggestions in Image Sharing Sites", which has been submitted to KASW 2008.

I am very pleased with this concrete outcome of our collaborative efforts, shortly after they've started. I look forward to the implementation and evaluation stages of this project and -- hopefully -- to many other papers and projects to come.

Course

My course on 'Selected Topics in Visual Information Retrieval' started very well and seems to be moving forward at a very high speed. Here are some highlights:

• I have a small but very qualified group of students: 10 in total, three of which are PhD students.
  
• They are all busy working on their implementation-oriented projects and the partial results (I've divided the project into milestones to better check their progress and help them along the way) are very encouraging.
  
• Due to limited classroom availability, the class meets in different rooms and buildings depending on the day, for a total of seven different classrooms. Talk about a great way to learn my way around campus... :-)
  
• Speaking of classrooms, an interesting story: before my first lecture I wanted to check if the classroom had projector, etc. I was told how to look up this information online. To my surprise, they have a system that contains that information and much more... Starting from the course page, when one clicks on the classroom number, a popup window shows information about the resources available in that room, and -- in the best German / Austrian tradition -- detailed quantitative information about the room's length, width, total area, number of seats, etc. Great!
  
• I'm using Moodle (in German!), which is a great CMS tool for online learning (that I was scheduled to test drive at FAU in the Fall anyway, so I'm enjoying the head start, even if some things have been learned by trial-and-error due to my notoriously limited knowledge of the German language).
  

Back to blogging...

I can't believe it's been almost a week since my last post... This is the longest I've been away from posting since this blog was created. 

The main reason is that things picked up a lot of speed here at Klagenfurt University, with my course in full motion, my first joint paper with colleagues from the Department, an invitation to go to Graz tomorrow, and many new developments, all very good and positive!

I'll blog about some of the many things that have happened over the past few days in separate posts. Stay tuned...

WIAMIS 2008: technical notes (Part 3)

And here is part 3 of the series... It refers to the second day of the conference. 

The day started with John R. Smith's very rich invited talk on "Unleashing video search". John presented an excellent talk, broad, deep, and current, supported by excellent graphics and demos. I liked (a lot!) when he said that all this incessant talk about bridging the semantic gap may lead to the trap of "building a bridge to nowhere". Maybe we should ask, said John (and I paraphrase): "What, exactly, is the destination?"

On a personal level, it was great to finally meet him in person (among other things, John contributed a chapter on "Video Indexing and Retrieval using MPEG-7" for the Handbook of Video Databases).

The next session (Oral Session II: Analysis II) had two papers on SenseCam-related work by the The Centre for Digital Video Processing (Dublin City University, Ireland), briefly commented in an earlier post.

After the lunch break, I attended the first and second parts of the Special Session on Multimedia Metadata Management and Retrieval (M3R). 

Between the two, I took a close look at the poster session for the day and was impressed by the work presented by Prof. Kai Uwe Barthel (FHTW-Berlin), whose demo is available online. It builds up on previous work (ImageSorter, available for download) and shows a lot of promise in terms of narrowing the infamous semantic gap.

Birthday (and cake!)

Today I celebrated my birthday (and my colleague Hermann Hellwagner's birthday) in style: with a delicious espresso macchiato and a birthday cake kindly baked by Martina Steinbacher (her recipe of the ultra-secret and world-famous Sachertorte). Scrumptious!

Later in the afternoon I gave my first lecture to a group of bright and interested students.

I want to thank everyone at Klagenfurt University for making this day a very pleasant day that I'll fondly remember for many years to come.

WIAMIS 2008: technical notes (Part 2)

And here is the second part of a series of postings on the technical side of WIAMIS 2008.  It refers to the afternoon portion of the first day. 

The first session (COST 292 – Towards Automated Robust Vision-based Surveillance) was chaired by Prof. Noel E. O'Connor, whose research team has an outstanding track record in video processing and related areas. During the conference I had the pleasure of meeting Prof. O'Connor and had a great time chatting with two other team members during our bus trip to Magdalensberg: Ciarán Ó Conaire and Aiden Doherty. They both presented papers during WIAMIS and could occasionally be spotted wearing their SenseCams. I commended Ciarán for posting MATLAB code for much of his research work for the benefit of others in the Computer Vision community.

In this session I also had the privilege of listening to my Portuguese colleague Paulo Lobato Correia present the paper "Traffic Sign Recognition Based on Pictogram Contours", by Carlos Filipe Paulo and Paulo Lobato Correia, which brought memories of the FAU Salient dataset.

After the coffee break we had a Special Session: RUSHES – Multimedia Indexing and Retrieval of Unedited Audio-Visual Footage, in which the paper "Describing Objects with Multiple Features for Visual Information Retrieval and Annotation", by Qianni Zhang and Ebroul Izquierdo, was presented. I met Prof. Izquierdo for the first time during WIAMIS and was delighted to see what his group is doing. In addition to the paper above, another paper/poster caught my attention: "Global-to-Local Oriented Rapid Scene Perception", by Le Dong and Ebroul Izquierdo. I plan to follow up with Prof. Izquierdo's group on both of them.

Fernando Pereira

Durante a WIAMIS 2008 tive a grande satisfação de conhecer pessoalmente o Prof. Fernando Pereira (Instituto Superior Técnico, Lisboa, Portugal), mundialmente famoso por seus muitos anos à frente de comitês MPEG e pelas renomadas publicações associadas, dentre elas "The MPEG-4 book".

Personalidade extrovertida e cativante, fez de nossas conversas (em português, é claro) alguns dos momentos mais agradáveis desta memorável conferência.

WIAMIS 2008: technical notes (Part 1)

As promised, this is the first part of a series of postings on the technical side of WIAMIS 2008 (before anyone starts thinking that the conference was all about socializing and visiting beautiful places. :-) 

It refers to the morning portion of the first day.  

The day started with Horst Bischof's excellent invited talk on "Robust Person Detection for Surveillance using Online", where he managed to engage everyone -- even those who are not directly working in this field -- with a great speech, fully supported by excellent slides, video clips, and demos. His take-home message was "Person detection in surveillance scenarios is feasible using online learning methods", but what struck me the most was his team's approach to conservative learning, which could be summarized as follows: "it is better to get fewer, but good, examples of persons".

The next session (Oral Session I: Analysis I) had four papers, three of which are very relevant and related to work that my students, collaborators, and I have been doing during the past few years:

• "A Comparative Study of Classification Techniques for Knowledge-Assisted Image Analysis", by G. T. Papadopoulos, K. Chandramouli, V. Mezaris, I. Kompatsiaris, E. Izquierdo, and M. G. Strintzis
• Follows up on the group's work reported recently in a rich and inspiring Eurasip JASP paper.
• "A Semantic Multimedia Analysis Approach Utilizing a Region Thesaurus and LSA", by Evaggelos Spyrou, Giorgos Tolias, Phivos Mylonas, and Yannis Avrithis
• Explores a combination of the (growingly popular) RSST segmentation algorithm, selected MPEG-7 descriptors, K-means clustering, and LSA in video documents.
• I'm curious to know where the next stages of their work (which plan to include context) will take them.
• "Exploiting Spatial Context in Image Region Labelling Using Fuzzy Constraint Reasoning", by Carsten Saathoff and Steffen Staab
• Learning and modeling spatial context is a hot topic. In this work (by a world-famous Semantic Web-oriented group) it is approached from a Fuzzy Constraint Reasoning perspective.

These papers (and associated references) have pointed me in directions that I had not considered before and given me many ideas that I hope to put into practice soon. 

If you're reading this blog and are interested in these topics (semantic image annotation and classification), please drop me an email or a comment on this entry.

WIAMIS 2008: group picture and farewell

WIAMIS 2008 was an extraordinarily successful event by every measure. Excellent talks and presentations on relevant contemporary topics, insightful observations from leading researchers and industry experts, and an overall atmosphere of friendly and focused discussions and interactions.

A group picture of the participants is now available on the conference official blog.

I learned a lot during the conference and would like to publicly thank Professors Laszlo Böszörmenyi and Hermann Hellwagner for their kind invitation.

Even though we've said our farewells and the conference is officially over, I still have a few technical posts related to WIAMIS 2008 "in the pipe". Stay tuned...

Multimedia Metadata Community

During WIAMIS 2008 I had the pleasure of meeting with the folks behind the Multimedia Metadata Community. I attended to most lectures of the 8th Workshop of the Multimedia Metadata Community (one of the special sessions within WIAMIS 2008) and joined the group discussions during the community meeting on Friday.

I thank Mathias Lux, Ralf Klamma, Michael Granitzer, Marc Spaniol, Romulus Grigoras, and Vincent Charvillat, among many others, for welcoming me into the community. I look forward to being part of some the great upcoming developments (events, projects, dissemination and collaboration initiatives) that the group outlined during the past few days.

WIAMIS 2008: photos

A set of photos taken on the first day of the conference (during the Welcome Reception) has been posted on flickr.

WIAMIS 2008 updates

WIAMIS 2008 started off in great style, with an excellent invited talk by Prof. Horst Bischof and many interesting technical presentations. At the end of a busy and productive day, all participants walked to beautiful Maria Loretto castle for the Welcome Reception.

The conference's blog has pictures and other updates.

Klagenfurt and WIAMIS

I've just arrived in beautiful Klagenfurt. Tomorrow I'll attend the first day of the WIAMIS conference. The conference program looks great! I look forward to learning from some of the world's experts and exchanging ideas with them on topics of common interest.