Students: for current available research topics, check the "Open Topics and Assignments" section.


Current Research Topics

My current focus very much lies in the domain of Document Image interpretation, and how raw data, extracted from document image processing techniques can be elevated to the rank of information through the combination of formal[13] and statistical learning techniques on the one hand, and the modeling of the context semantics on the other hand. 

General Interests and Previous Subjects

Reproducible Research and Performance Analysis

Part of this work is very tightly related with the DARE paradigm and the associated DAE platform and its follow-up project DIA-Tribe.

Digital Humanities

Various on and off going interactions with the ANRT Typography research lab in Nancy, in trying to figure out how image analysis can help capturing typographic information in Renaissance documents (Re-Typograph). Partially in collaboration with Boise State University.

Graphical Document Image Analysis

From 2000 to 2009, I've been involved in graphical document analysis mainly focusing on extracting robust features[12] for indexing[11] and recognition[10] of graphical document collections[9].

Visual Servoing

During 1998-2000 I took a brief excursion into the domain of visual servoing and tracking. Besides a significant part of technology transfer, main scientific contributions related to metric and non metric visual stereoscopic robot control by addressing uncalibrated stereo control through integration of epipolar constraints[5,6] into previous results[7]. Other contributions consisted in work on fully un-calibrated stereo control methods[8].

Image Recognition and Indexing

This is work that was mainly done during my PhD. (1994-1998). It concerns object recognition in real-world images from structural data, indexed using affine and projective quasi-invariants indexing keys. Initially only considering image contours[1], it was extended to geometrically coherent combinations of contours and interest patches[2,3]. It may actually give a key to integrating geometric coherence to bag-of-word methods. It was successfully tested in an industrial environment for the identification of mechanical parts of car engines.

One of the more fundamental parts of this work has been the study of the inherent computational complexity related to content-based image indexing[4]. This has been one of the first published works noticing that indexing multimedia data imperatively needs to take into account computational complexity related to high dimensional metric spaces


Please refer here for a full list of publications.

[1] Rapid Object Indexing and Recognition Using Enhanced Geometric Hashing, B. Lamiroy and P. Gros in Proceedings of the 4th European Conference on Computer Vision, Cambridge, United Kingdom, Vol. 1, pp. 59-70, April 1996.

[2] Combining Local Recognition Methods for Better Image Recognition, B. Lamiroy, P. Gros and S. Picard in the Eleventh British Machine Vision Conference, Bristol, United Kingdom, 11-14 September 2000.

[3] An Image Oriented CAD Approach, C. Schmid, Ph. Bobet, B. Lamiroy et R. Mohr in Proceedings of the ECCV'96 International Workshop on Object Representation in Computer Vision, Cambridge, United Kingdom, pp. 221-245, April 1996.

[4] Object Indexing is a Complex Matter, B. Lamiroy and P. Gros in Proceedings of the 10th Scandinavian Conference on Image Analysis, Lappeenranta, Finland, Vol. I, pp. 277-283, June 1997

[5] Controlling Robots with Two Cameras: How to Do it Properly, B. Lamiroy, B. Espiau, N. Andreff et R. Horaud, in IEEE International Conference on Robotics and Automation, San Francisco, United States of America, April 24-28 2000.

[6] What Metric Stereo Can Do for Visual Servoing, B. Lamiroy, C. Puget and R. Horaud, in IEEE/RSJ International Conference on Intelligent Robots and Systems, Kagawa University, Takamatsu, Japan, October 30 - November 5, 2000.

[7] Object Pose: The Link between Weak Perspective, Paraperspective and Full Perspective, R. Horaud, F. Dornaika, B. Lamiroy and S. Christy in International Journal of Computer Vision, No. 2, pp. 173-189, 1997.

[8] Visual Control Using Projective Kinematics, A. Ruf, F. Martin, B. Lamiroy and R. Horaud in John Hollerbach and Dan Koditschek (Eds.) Robotics Research: the Ninth International Symposium, Springer-Verlag, London, 2000.

[9] Scan-to-XML: Using Software Component Algebra for Intelligent Document Generation, B. Lamiroy, L. Najman, in 4th International Workshop on Graphics Recognition - Algorithms and Applications - GREC'2002, Oct., 2002, Kingston, Ontario, Canada, col. Lecture Notes in Computer Science, vol. 2390, pp. 211-221.

[10] Automatic Adjacency Grammar Generation from User Drawn Sketches, J. Mas Romeu, B. Lamiroy, G. Sánchez, J. Lladós, in 18th International Conference on Pattern Recognition - ICPR 2006, Aug., 2006, Hong Kong/Chine, vol. 2, pp. 1026 - 1029.

[11] A Few Steps Towards On-the-Fly Symbol Recognition with Relevance Feedback, J. Rendek, B. Lamiroy, K. Tombre, in 7th IAPR Workshop on Document Analysis Systems - DAS 2006, Feb., 2006, Nelson/New Zealand, col. Lecture Notes in Computer Science, vol. 3872, pp. 604 - 615.

[12] Robust and Precise Circular Arc Detection, B. Lamiroy, Y Guebbas, in 8th International Workshop, GREC 2009, Jul., 2009, La Rochelle, France, col. Lecture Notes in Computer Science, vol. 6020, pp. 49-60.

[13] Inductive Logic Programming for Symbol Recognition, S. K.C., B. Lamiroy, J.-P. Ropers in Tenth International Conference on Document Analysis and Recognition - ICDAR'2009, Jul. 26, 2009, Barcelona, Spain, pp. 1330 - 1334.