September 17, 2012

Forenoon session (10:00-13:00) (Including 15 minutes coffee break during 11:15-11:30)

Afternoon session (16:00-19:00) (Including 15 minutes coffee break during 17:15-17:30)

Tutorial A

Signature Verification - Forensic Examiners` Perception and Solutions for Off-Line and On-Line Signatures

(by Marcus Liwicki. Micahel Blumenstein and Elisa van den Heuvel)

In this tutorial we will give insights into the daily work to be performed by professional Forensic Handwriting Examiners (FHEs) and give examples for their case-work. Furthermore, we will give an overview on the work done in the pattern recognition area and a comparison of state-of-the-art automatic verification systems. In the second half we will introduce two automatic approaches for off-line and on-line signature verification with more detail and explain all the processing steps. Additionally, we will provide parts of these systems as a starting point for the development of identification systems for the participants.

Tutorial B

15 Years of Handwritten Word Spotting: Achievements and Remaining Challenges

(By Alicia Fornés, Marçal Rusiñol and Volkmar Frinken)

The problem of word spotting can be defined as the task of locating the regions within a document image which are likely to contain an instance of a given queried word without explicitly recognizing it. Word spotting techniques are particularly interesting to provide accessibility to document collections in which the performance of OCR engines is still poor. One of the most common scenarios is the use of word spotting techniques in handwritten documents.
The first works on handwritten word spotting were published 15 years ago. In this tutorial we aim to review the word spotting techniques. Starting from well-known references from the early years, we will travel through the key word spotting techniques used in the handwriting recognition field until the most recent publications while pointing the achievements and the remaining challenges. The agenda is organized in self-contained lectures addressed by three specialists in the topic of handwritten word spotting.

Tutorial C

Handwriting recognition - Indic scripts

(by A G Ramakrishnan, Suresh Sundaram, Sitaram Ramachandrula and Hariharan Ravishankar)

The tutorial will address the state-of-the-art technology in online and offline handwriting recognition (HWR), with a focus on Indic scripts. The uniqueness of each script, together with the challenges posed in the design of online and offline recognition systems, will be emphasized. The tutorial will highlight how technologies used for non-Indic scripts have been tailored for Indic script recognition.
The tutorial will mainly address young, budding researchers and give them an overview of the general fields of online and offline handwriting recognition, with particular emphasis on challenges with respect to Indic language handwriting.
There are 14 Indic scripts and there is a huge untapped potential for Indian population to access Information Technology through Indian languages. Handwriting being a natural interface to computers, especially with the huge success of iPad, camera mobile phones, etc., handwriting recognition of Indian scripts is a topic of emerging importance and considerable interest for the industry. Indian scripts offer a huge area for research, where hitherto, not much work has been done.

Tutorial D

Development of HMM based Handwritten Text/Word Recognition Systems – Conception, Implementation and Evaluation

(by Volker Märgner, MahdiHamdani and Haikal El Abed)

The aim of the tutorial “Development of HMM based Handwritten Text/Word Recognition Systems – Conception, Implementation and Evaluation” is to lay the foundations and to encourage further discussions on the development of pattern recognition systems, especially for the recognition of handwritten text. Researchers and practitioners working in the field of pattern recognition will be introduced to handwritten text recognition systems in general, different state-of-the-art approaches, steps of a system development process and techniques to evaluate and improve recognition quality. The objective of this tutorial is to provide a basis for researchers to develop an efficient recognition system and to improve the quality and performance of the developed system. The presented methods include detailed analysis of a recognition system, relation between structured data and performance of systems, reject/combination strategies, and post-processing approaches. After this tutorial participants should be able to design their own recognition system or improve an existing one.

Tutorial E

Information retrieval from handwritten documents

(By U. Porwal, V. Govindaraju and A. Bhardwaj)

This tutorial aims to provide a basic understanding of existing techniques for information retrieval from handwritten documents. The scope of the tutorial is to familiarize researchers from various backgrounds with this key topic and help them apply such techniques to related research problems as well as real world applications. The first part of the tutorial would focus on introductory concepts of information retrieval, handwriting recognition and image processing. Text based retrieval models involving noisy text retrieval as well as adaptation of traditional retrieval models will also be described. The second part of the tutorial would survey the existing IR toolkits available to researchers and discuss important concepts from keyword spotting. Topics from recognition based as well as recognition free keyword spotting would be covered in detail. Content based retrieval of handwriting in form of signature based document retrieval, writer style retrieval and temporal document retrieval would also be discussed.