IEEE ICPRS 2027

Explainable Document Analysis for Domain Experts: Combining Deep Learning with Rule-Based Methods

Video understanding has advanced dramatically over the past fifteen years, from hand-crafted visual features and concept detectors to deep neural With the advent of end-to-end trainable, data-driven deep learning methods, a large number of classical document analysis tasks, such as layout analysis, OCR, handwriting recognition, and information retrieval underwent significant improvements in accuracy. However, at the same time, the "black box" effect of these methods remains challenging for domain experts, who rely on automatic document analysis in their profession, and who have to take responsibility for their observations and conclusions. In this talk, I will present several case studies that highlight benefits of combining deep learning with rule-based methods, with a view to explainability, accuracy, and human annotation effort.

Andreas Fischer is Full Professor at the University of Fribourg, Switzerland, in the Department of Informatics, where he leads the research group on AI for the Benefit of Human Experts (AIBEX). He obtained his PhD at the University of Bern, conducted postdoctoral research at Concordia University and Polytechnique Montréal, and held, prior to his appointment, a combined position in Fribourg as Lecturer at the University of Fribourg and as Professor at the School of Engineering and Architecture of Fribourg. Andreas Fischer's research interests include deep learning, pattern recognition, document analysis, handwriting recognition, digital humanities, and digital pathology. He has published over 120 peer-reviewed articles in international journals and conference proceedings on these topics. Andreas Fischer is a member of the governing board of the International Association of Pattern Recognition (IAPR), where he represents Switzerland, and Chair of the IAPR technical committee on reading systems (TC11).