2020, Stauffer, Michael, Fischer, Andreas, Riesen, Kaspar

2019, Ameri, Mohammad Reza, Stauffer, Michael, Riesen, Kaspar, Bui, Tien Dai, Fischer, Andreas, Fischer, Andreas

Keyword spotting enables content-based retrieval of scanned historical manuscripts using search terms, which, in turn, facilitates the indexation in digital libraries. Recent approaches include graph-based representations that capture the complex structure of handwriting. However, the high representational power of graphs comes at the cost of high computational complexity for graph matching. In this article, we investigate the potential of Hausdorff edit distance (HED) for keyword spotting. It is an efficient quadratic-time approximation of the graph edit distance. In a comprehensive experimental evaluation with four types of handwriting graphs and four benchmark datasets (George Washington, Parzival, Botany, and Alvermann Konzilsprotokolle), we demonstrate a strong performance of the proposed HED-based method when compared with the state of the art, both, in terms of precision and speed.

2018, Stauffer, Michael, Fischer, Andreas, Riesen, Kaspar

In the last decades historical handwritten documents have become increasingly available in digital form. Yet, the accessibility to these documents with respect to browsing and searching remained limited as full automatic transcription is often not possible or not sufficiently accurate. This paper proposes a novel reliable approach for template-based keyword spotting in historical handwritten documents. In particular, our framework makes use of different graph representations for segmented word images and a sophisticated matching procedure. Moreover, we extend our method to a spotting ensemble. In an exhaustive experimental evaluation on four widely used benchmark datasets we show that the proposed approach is able to keep up or even outperform several state-of-the-art methods for template- and learning-based keyword spotting.

2019, Stauffer, Michael, Maergner, Paul, Fischer, Andreas, Riesen, Kaspar

Due to the high availability and applicability, handwritten signatures are an eminent biometric authentication measure in our life. To mitigate the risk of a potential misuse, automatic signature verification tries to distinguish between genuine and forged signatures. Most of the available signature verification approaches make use of vectorial rather than graph-based representations of the handwriting. This is rather surprising as graphs offer some inherent advantages. Graphs are, for instance, able to directly adapt their size and structure to the size and complexity of the respective handwritten entities. Moreover, several fast graph matching algorithms have been proposed recently that allow to employ graphs also in domains with large amounts of data. The present paper proposes to use different graph embedding approaches in conjunction with a recent graph-based signature verification framework. That is, signature graphs are not directly matched with each other, but first compared with a set of predefined prototype graphs, in order to obtain a dissimilarity representation. In an experimental evaluation, we employ the proposed method on two widely used benchmark datasets. On both datasets, we empirically confirm that the learning-free graph embedding outperforms state-of-the-art methods with respect to both accuracy and runtime.

2019, Stauffer, Michael, Maergner, Paul, Fischer, Andreas, Riesen, Kaspar, Conte, Donatello, Ramel, Jean-Yves, Foggia, Pasquale

In contrast to statistical representations, graphs offer some inherent advantages when it comes to handwriting representation. That is, graphs are able to adapt their size and structure to the individual handwriting and represent binary relationships that might exist within the handwriting. We observe an increasing number of graph-based keyword spotting frameworks in the last years. In general, keyword spotting allows to retrieve instances of an arbitrary query in documents. It is common practice to optimise keyword spotting frameworks for each document individually, and thus, the overall generalisability remains somehow questionable. In this paper, we focus on this question by conducting a cross-evaluation experiment on four handwritten historical documents. We observe a direct relationship between parameter settings and the actual handwriting. We also propose different ensemble strategies that allow to keep up with individually optimised systems without a priori knowledge of a certain manuscript. Such a system can potentially be applied to new documents without prior optimisation.

2019, Stauffer, Michael, Maergner, Paul, Fischer, Andreas, Ingold, Rolf, Riesen, Kaspar

In recent years, different approaches for handwriting recognition that are based on graph representations have been proposed (e.g. graph-based keyword spotting or signature verification). This trend is mostly due to the availability of novel fast graph matching algorithms, as well as the inherent flexibility and expressivity of graph data structures when compared to vectorial representations. That is, graphs are able to directly adapt their size and structure to the size and complexity of the respective handwritten entities. However, the vast majority of the proposed approaches match the graphs from a global perspective only. In the present paper, we propose to match the underlying graphs from different local perspectives and combine the resulting assignments by means of Dynamic Time Warping. Moreover, we show that the proposed approach can be readily combined with global matchings. In an experimental evaluation, we employ the novel method in a signature verification scenario on two widely used benchmark datasets. On both datasets, we empirically confirm that the proposed approach outperforms state-of-the-art methods with respect to both accuracy and runtime.

2018, Stauffer, Michael, Fischer, Andreas, Riesen, Kaspar

Scanned handwritten historical documents are often not well accessible due to the limited feasibility of automatic full transcriptions. Thus, Keyword Spotting (KWS) has been proposed as an alternative to retrieve arbitrary query words from this kind of documents. In the present paper, word images are represented by means of graphs. That is, a graph is used to represent the inherent topological characteristics of handwriting. The actual keyword spotting is then based on matching a query graph with all document graphs. In particular, we make use of a fast graph matching algorithm that considers the contextual substructure of nodes. The motivation for this inclusion of node context is to increase the overall KWS accuracy. In an experimental evaluation on four historical documents, we show that the proposed procedure clearly outperforms diverse other template-based reference systems. Moreover, our novel framework keeps up or even outperforms many state-of-the-art learning-based KWS approaches.