[TC11 Newsletter/Nov-2002] Date: 12/06/2002 07:55 PM Dear colleagues, many of you will be working on your paper for ICDAR 2003: good luck! Note: deadline extension message: >> In response to many requests, we are pleased to announce an >> extension of the deadline for paper submission to ICDAR2003 >> to Midnight, Sunday January 5th 2003 PST, 8am Monday >> January 6th 2003 GMT. This will be the final deadline and >> there will be no further extensions, otherwise the review >> process will not be able to proceed in good time. >> Andy Downton, Conference Chair >> Mike Fairhurst, Conference Co-Chair >> Apostolos Antonacopoulos, Publications Chair Furthermore, I have three exciting news items for you: * 1. As discussed at the previous ICPR, continuity and modernization are important aspects of governance within the IAPR Technical Committees at large, including TC-11. A solution has been found by the installment of the function of Vice Chair, who will assist the current TC-11 chair and who will be able to get accustomed with the tasks involved. I am now very happy to announce a new Vice Chair for TC-11: Ms. Jianying Hu, who presented a convincing pamphlet for the directions of research in TC-11 [see attachment]. Already, I would like to thank Ms. Hu for taking up this responsibility. From our new leader: > It is my pleasure to re-appoint you officially > as chairman of IAPR TC11 on Reading Systems. > Based on your recommendation I would like to appoint also > Jianying Hu as vice chair for TC-11. > IAPR expects technical committees to promote research activities > in their respective fields of specialization. Usually this is done by > organizing workshops and meetings. Another activity is to create and > maintain a web-page and a membership list to enhance its visibility > and attract further members through the web. Please continue your > good activities for TC11 in the above sense. > Walter Kropatsch, > IAPR 1st vice president * 2. Ms. Katrin Franke was so nice to provide her report on the IWFHR in Niagara on the Lake, this year. Handwriting research is alive and kicking, so it seems! [see attachment] * 3. There is more movement ahead, in the area of camera-based text and document recognition! The International Journal on Document Analysis and Recognition is issuing a call for papers on this topic [see PDF attachment]. Finally, if there are news items, don't hesitate to forward them to me, I will be happy to collect and broadcast them. Best regards, Lambert Schomaker Grote Kruisstraat 2/1, 9712 TS Groningen, The Netherlands Tel: +31-50-363-7908 / Fax: +31-50-363-6687 TC11 Vice Chair Plans Jianying Hu While the traditional focus of the field of document image analysis (DIA) started out with the analysis and conversion of paper documents, it is clear that the meaning of "document" has been expanding dramatically over the last ten years. Being able to adapt to new ways documents are being created and consumed in an ever more technical savvy society is crucial to maintaining the relevance and vigor of this field. I believe TC11 could and should play a major role in promoting research in and attracting young researchers to the new research areas vital to the growth of this community. To name a few: 1. Web Document Analysis (WDA). With the ever-increasing use of the Web, a growing number of documents are published and accessed on-line. Although the development of XML and the new efforts on semantic web aim to improve the machine-readability of web documents, it is not going to eliminate the need for content analysis, especially for the kind of web documents created as web publications (vs. services) where visual appearance is critical. Such content analysis is crucial for many application including information extraction, mining, summarization and content re-purposing for mobile access, and this is where DIA can play a unique role. There has been growing attention to WDA within the DIA community, especially following the first WDA workshop held in conjunction with ICDAR'01. Many interesting areas have emerged besides content analysis mentioned above, including utilizing and mining multilingual documents, web page classification, and human interactive proofs (HIP). TC11 could help nurturing this new field through active support and promotion of the WDA workshop series, reaching out to and engaging researchers in IR and WWW communities (e.g., in the form of tutorials, invited speeches, etc.), and improving our own visibility in those communities. 2. Electronic ink/Mobile user interfaces/annotation. The success of the latest IWFHR indicates that interest in handwriting recognition is still quite high. One important new drive for electronic ink processing is the increasing capacity and popularity of hand held computers of all forms. Issues in pen based interface design is attracting renewed attention, especially in the context of multi-modal user interfaces for mobile devices, and collaboration and document annotation using such devices. InkXML is in the works and TC11 can play an active role to publicize this effort and the related new research activities. 4. Vintage documents/Digital Library. Another area that has attracted increasing attention from this field lately is the analysis of vintage documents, related to the Digital Library effort. This offers unique opportunities as well as challenges for DIA, but has not been well explored so far by this community. Again, TC11 could help by bringing more awareness to this area through the official website. September 2002 The historic town Niagara-on-the-Lake, at the mouth of the Niagara River, was hosting the 8th International Workshop on Frontiers in Handwriting Recognition from 6th ?8th August 2002. The main objective of this workshop is to stimulate contacts between researchers and practitioners, and to motivate in-depth discussions on research results and frontiers in the field of handwriting recognition and analysis. Even after more than thirty years of research in handwriting recognition, with systems operating in actual practice, and products being available on the consumer market, many scientific challenges will remain to be solved. This year's workshop sessions were focused on learning methods, classifier design, multiple classifiers, on-line/ off-line handwriting recognition, word recognition, pen computing, document applications, postal applications, signature verification and writer identification as well as forensic document examination. Since 1990 the workshop is being organized as satellite event to the International Conference on Pattern Recognition (ICPR), which was hosted in Que'bec City, Canada, this year. Remembering the discussions for a suitable IWFHR-workshop location four years ago, there were strong arguments for Niagara-on-the-Lake with its natural beauty and tourist's attractions. Moreover, the region offers a deep insight into the culture and history of North America. Niagara-on-the-Lake is being always a place where Canadians and US citizens meet. So, it seems to be also symbolic that the workshop was co-chaired by S.-N. Srihari from Center of Excellence for Document Analysis and Recognition (CEDAR), State University of New York at Buffalo, United States, and M. Cheriet from E'cole de Technologie Supe'rieure, Canada. Of course, not only the magnificent workshop location but also the excellent scientific program attracted an international audience. All together 144 university members as well as representatives of governmental organizations and industrial entities coming from 21 countries attended the workshop. During three memorable days participants followed 51 oral presentations and three panels. In the course of three poster sessions with 36 presentations the possibilities for individual discussions with the poster presenters was given. Among the papers that were accepted for presentations, the international program committee selected the two best papers on the basis of a fair evaluation process. In the category senior research G. Seni from Motorola Inc. was honored for his outstanding research and presentation titled 'TreadMill Ink ? Enabling Continuous pen Input on Small Devices'. Best paper in the category junior researcher was devoted to C. Bahlmann from the University of Freiburg, Germany, for his work and presentation on 'On-line Handwriting recognition with Support Vector machines ? A kernel approach.' IWFHR-workshops are always a platform to recover challenging research topics. This year, the panels sessions, in particular panels on pen computing and forensic document examination, pointed to new directions and drawn a future of applications employing handwriting recognition and analysis. In the course of the pen-computing panel, presenters agreed that the time for a broad usage of pens for computer-interaction has come. First consumer products, such as mobile phones, handheld and mobile computers are on the market. Moreover, next generation of pen computers is already prototypically realized. However, there is still the great demand in the standardization of pen-input data. So, it was an excellent opportunity to present the brand-new InkXML-specification in the evening of the first workshop day. InkXML is promoted by researchers and practitioners from IBM Corp., Intel Corp., Motorola Inc. and the International Unipen Foundation (iUF). So far, the InkXML draft was officially contributed to the W3C - Multimodal Interaction working group, and has recently been posted in the public archives. During the panel on forensic document examination practitioners from forensics, in particular from US Secret Service, Federal Investigation Bureau (FBI), Immigration and Naturalization Service (INS)} as well as from Canada Custom stated their expectations from research in the field of handwriting recognition and analysis. Areas of interest are identification of writers based on static samples of handwriting, relations from psycho-motor handwriting production to the visual appearances of handwriting as well as recovering of language and writing system specific patterns in handwritten products. T. Moran, from US Secret Service, gave impressive presentation on state-of-the-art technologies in computer-based forensic handwriting analysis as used in forensic labs. However, the dramatic increase of white-collar crime derives a great demand of high sophisticated approaches to computer-based forensic handwriting examination. The soaring interest and the expected benefit form research in the field of handwriting recognition and analysis is also reflected by the number and superiority of workshop sponsors that are (in alphabetical order) A2ia Corp, Hitachi Ltd., International Association of Pattern Recognition (IAPR), Microsoft Corporation, Motorola Inc., Siemens AG and United States Postal Service. Nevertheless, their support and the marvelous enthusiasm of the many colleagues from Center of Excellence for Document Analysis and Recognition (CEDAR) guaranteed a successful workshop. On the last evening of the workshop, participants gathered in the exquisitely designed Cha^teau des Charmes. After following a guided walking tour through the harvest room and underground barrel cellars, participants enjoyed the vine and the beautiful view over the vineyards. Later that night, C. Suen, honorary chair of this year^(1)s workshop and initiator of the IWFHR workshop series in 1990, raised his glass to praise colleagues and friends who promoted IWFHR series so far. In recognition for their distinguished contributions, the organizers of the past workshops were honored with one of the last copies of the 1990-workshop-proceeding becoming 'antique' and in great demand in the meantime. The IWFHR-community also wants to thank previous workshop organizers and their teams, being in 1991 S. Impedovo and J.C. Simon, in 1993 S.-N. Srihari, 1994 J.-F. Wang, 1996 A.C. Downtown and S. Impedovo, in 1998 J.-H. Kim, in 2000 L. Schomaker and in 2002 S.-N. Srihari and M. Cheriet. The next IWFHR-workshop will be organized by H. Fujisawa, from Hitachi Ltd.. >From October, 26th-28th, 2004 we will meet in Kokubunji/Tokyo, Japan, for the first time. We look forward to this next highlight and the world-famous hospitality of the Japanese. Meanwhile we will stick to our research and developments demanding our attention and creativity. However, keep in mind, next deadline, 15th March 2004, is coming faster than expected. For all those who wish to personally interchange with colleagues in 2003, International Conference on Document Analysis and Recognition (ICDAR - http://www.essex.ac.uk/ese/icdar2003/) provides an excellent platform as well. The conference will be hosted in Edinburgh, Scotland from 23th -26th August 2003. I look forward to meet and discuss with you virtually over the Internet or on a personal base during one of the next conferences. Katrin Franke