Prof. Dr.-Ing. H. Siegfried Stiehl

2015-2019 PI of Scientific Service Project Z03 "Image Processing
Methods for Determining Visual Manuscript and Character Features" in
Collaborative Research Center (Sonderforschungsbereich) 950
"Manuscript Cultures in Asia, Africa and Europe" (DFG)"

https://www.manuscript-cultures.uni-hamburg.de/Poster/Z03_A4_P2_eng.pdf

Team:

Märgner, Volker, Dr.-Ing., Co-PI
Academic Director ret.
Institute for Communications Technology, TU Braunschweig
Tel.: +49 531 391 2483
Fax: +49 531 391 8218
v.maergner [at] tu-bs.de
www.ifn.ing.tu-bs.de

Konidaris, Thomas, PhD
PostDoc (Word Spotting)
Tel: +49 40 42838 - 9394
Fax: +49 40 42838 - 4899
thomas.konidaris [at] uni-hamburg.de

Mohammed, Hussein Adnan, M.Sc.
Doctoral Student (Handwriting Analysis and Writer Identification)
Tel: +49 40 42838 - 9397
Fax: +49 40 42838 - 4899
hussein.adnan.mohammed [at] uni-hamburg.de

Publikationen:

Efficient Keypoint Reduction for Document Image Matching (T. Konidaris, V. Märgner, H.A. Mohammed, H.S. Stieh), In: Proc. 8th Int. Conf. on Pattern Recognition Applications and Methods, Prague, February 19-21, 2019 (pdf).

Writer identification in historical manuscripts: Analysis and optimization of a classier with an easy to use implementation for scholars from the humanities (H. Mohammed, V. Maergner, and H. S. Stiehl), In: 16th IAPR International Conference on Frontiers in Handwriting Recognition (ICFHR), page to appear. IEEE, Aug 2018. (pdf)

NormalisedLocal Naive Bayes Nearest-Neighbour Classifier for Offline Writer Identification (Hussein Mohammed, Volker Märger, Thomas Konidaris, H. Siegfried Stiehl), In: Document Analysis and Recognition (ICDAR), 14th International Conference on. IEEE, 2017, Kyoto, Japan, pp. 1013-1018. (pdf)

OpenX for Interdisciplinary Computational Manuscript Research (Pre-Conference Workshop)

June 12 - 13, 2018

CSMC, The Centre for the Study of Manuscript Cultures
University of Hamburg, Warburgstraße 26,
20354 Hamburg, Germany 

Directors:

H. Siegfried Stiehl, Prof. Dr.-Ing. & Vinodh Rajan Sampath, PhD

Technician: Dieter Jessen

Affiliated Z03 Team Members:

Thomas Konidaris, PhD, Hussein Adnan Mohammed, MSc

MS/BS Student Members: 

Tim Christopher Hahn (M): Computational Analysis of Charactêres in Digitized Manuscripts: A Proof of Concept (March 2019)

(Joint Supervision with Dr. Michael Kohs, SFB 950, Sub-Project B08)

Christoffer Kassens (M): Gabor Wavelets-Based Detection and Description of Repetitive Line-Like Texture in Digitised Documents: Theory and Experiments

Parth Sarthi Pandey (M): Detection of Visual Structure in Magia Figurate Manuscripts

Marlon Kramer (B): SIFT/FAST Performance Comparison

Nico Noster (B): Visualizing SIFT and FAST based workflow

iXMan_Lab is an acronym for interactive eXploration of Manuscripts Laboratory. The underlying motto for our laboratory - funded by DFG via SFB 950 "Manuscript Cultures in Asia, Africa and Europe"  - is to jointly develop theory-driven concepts, computational paradigms, experimental designs, and workflow-driven prototypes which they can utilize in their day-to-day research activities. (pdf)

Even though primarily situated within the Department of Informatics, the lab is uniquely placed within the Centre for Manuscript Studies (CSMC) as well through its ability to web-interact with various scholars from selected pilot sub-projects of SFB 950 – Manuscript Cultures in Asia and Africa. As a consequence, the interdisciplinary approach behind our laboratory adopts the facilitation of a joint requirement analysis in the light of design thinking along with a joint experimentation given computational methods for manuscript analysis through close interaction between scholars from Manuscript Studies and Informatics. In a broader context, the laboratory is embedded in the open science paradigm, e.g. open data initiatives and open source toolboxes. 

The lab provides an infrastructure, or even platform, for interdisciplinary teams by utilizing a multi-touch table environment with high-performance computing equipment as a medium for a two-fold aim: First, experimentally designing a manageable processing chain based on computational vision methods for analyzing digitized manuscripts and, second, freezing-in a validated (or even evaluated or benchmarked) processing chain by consensus in order to deliver a useful tool for a broad range of users. In terms of hardware capabilities, the laboratory currently has a custom built 65-inch multi-touch table (MTT) supported by a multi-core gaming engine. Thus our laboratory is completely equipped for running GPU-accelerated image processing and analysis algorithms, and if necessary, running deep learning methods as well. 

Currently, the main focus of our iXMan_Lab is set on the further development of the Advanced Manuscript Analysis Portal (AMAP) equipped with an intuitive interaction paradigm in the context of a multi-touch table. Hence our iXMan_Lab  will allow interdisciplinary teams to intuitively customize various advanced image processing and analysis methods as processing chains through visual programming, touch-based interaction on a large (65") scale and joint experimentation with digitized manuscripts from selected SFB sub-projects.

Publications:

Advanced Manuscript Analysis Portal (AMAP): An Interactive Visual Language Environment for Manuscript Studies (Vinodh Rajan Sampath, H. Siegfried Stiehl), In: Proceedings of Digital Humanities Conference 2019 (Annual International Conference of the Alliance of Digital Humanities Organizations)", July 9-12, 2019, Utrecht (to be published).

From Eye-to-Eye to Hand-in-Hand: Collaborative Solution Building in Interdisciplinary Manuscript Research (Rajan, V. & Stiehl, H. S., In: Burghardt, M. & Müller-Birn, C. (Hrsg.), INF-DH-2018. Bonn: Gesellschaft für Informatik e.V. (pdf)

Bringing Paleography to the Table: Developing an Interactive Manuscript Exploration System for Large Multi-Touch Devices (Vinodh Rajan Sampath, H. Siegfried Stiehl), In: Document Analysis Systems (DAS), 13th IAPR International Workshop, 2018, Vienna, Austria. (pdf)

Academic Education

Academic Career Milestones