Search

Nowadays, the accessibility of digitized historical documents is extremely important to facilitate fast and efficient retrieval of historical information and knowledge extraction from such data. To provide such functionality, it is necessary to convert document images into plain text using optical character recognition (OCR). Many OCR related methods and tools have been proposed, however, they are often too complicated for a standard user, some important parts are missing or they are not available in free versions. Therefore, this paper describes a complex and flexible web framework for historical document manipulation and analysis with the main focus on OCR. The framework contains eight modules to facilitate three main tasks: image pre-processing and segmentation, creation of data for OCR model training and the OCR itself. This framework is freely available for non commercial purposes. We have experimentally evaluated this framework on real data and we have shown that this system is efficient and can save human labour in the process of annotated data preparation. Moreover, we have reached state-of-the-art OCR results.

As the number of digitized historical documents has increased rapidly during the last a few decades, it is necessary to provide efficient methods of information retrieval and knowledge extraction to make the data accessible. Such methods are dependent on optical character recognition (OCR) which converts the document images into textual representations. Nowadays, OCR methods are often not adapted to the historical domain; moreover, they usually need a significant amount of annotated documents. Therefore, this paper introduces a set of methods that allows performing an OCR on historical document images using only a small amount of real, manually annotated training data. The presented complete OCR system includes two main tasks: page layout analysis including text block and line segmentation and OCR. Our segmentation methods are based on fully convolutional networks, and the OCR approach utilizes recurrent neural networks. Both approaches are state of the art in the relevant fields. We have created a novel real dataset for OCR from Porta fontium portal. This corpus is freely available for research, and all proposed methods are evaluated on these data. We show that both the segmentation and OCR tasks are feasible with only a few annotated real data samples. The experiments aim at determining the best way how to achieve good performance with the given small set of data. We also demonstrate that obtained scores are comparable or even better than the scores of several state-of-the-art systems. To sum up, this paper shows a way how to create an efficient OCR system for historical documents with a need for only a little annotated training data.

Článek se zabývá detekcí, klasifikací a rozpoznáváním toponym v ručně kreslených historických katastrálních mapách. Toponyma jsou místní jména měst, vesnic a dalších míst, jako jsou řeky, lesy atd. Extrahovaná toponyma se používají jako klíčová slova v systému vyhledávání informací, které umožňuje inteligentní a efektivní vyhledávání v historických mapových sbírkách. Vytvořili jsme novou datovou sadu, která je volně dostupná pro výzkumné účely. Poté navrhujeme nový přístup ke klasifikaci toponym založený na deskriptoru KAZE. Dále porovnáváme a vyhodnocujeme několik nejmodernějších metod pro detekci na naší úloze detekce toponym. Na závěr prezentujeme výsledky rozpoznávání textu toponym pomocí populárního enginu Tesseract. The paper deals with detection, classification and recognition of toponyms in hand-drawn historical cadastral maps. Toponyms are local names of towns, villages and landscape features such as rivers, forests etc. The detected and recognized toponyms are utilized as keywords in an information retrieval system that allows intelligent and efficient searching in historical map collections. We create a novel annotated dataset that is freely available for research and educational purposes. Then, we propose a novel approach for toponym classification based on KAZE descriptor. Next we compare and evaluate several state-of-the-art methods for text and object detection on our toponym detection task. We further show the results of toponym text recognition using popular Tesseract engine.

In this paper we exploit cross-lingual models to enable dialogue act recognition for specific tasks with a small number of annotations. We design a transfer learning approach for dialogue act recognition and validate it on two different target languages and domains. We compute dialogue turn embeddings with both a CNN and multi-head self-attention model and show that the best results are obtained by combining all sources of transferred information. We further demonstrate that the proposed methods significantly outperform related cross-lingual DA recognition approaches., Comment: Accepted for 17th International Conference on Artificial Intelligence Applications and Innovations (AIAI 2021), 25-27 June

This paper presents a set of methods for detection of important features in historical cadastral maps. The goal is to allow a seamless connection of the maps based on such features. The connection is very important so that the maps can be presented online and utilized easily. To the best of our knowledge, this is the first attempt to solve this task fully automatically. Compared to the manual annotation which is very time-consuming we can significantly reduce the costs and provide comparable or even better results.

Automatic writer identification is a common problem in document analysis. State-of-the-art methods typically focus on the feature extraction step with traditional or deep-learning-based techniques. In retrieval problems, re-ranking is a commonly used technique to improve the results. Re-ranking refines an initial ranking result by using the knowledge contained in the ranked result, e. g., by exploiting nearest neighbor relations. To the best of our knowledge, re-ranking has not been used for writer identification/retrieval. A possible reason might be that publicly available benchmark datasets contain only few samples per writer which makes a re-ranking less promising. We show that a re-ranking step based on k-reciprocal nearest neighbor relationships is advantageous for writer identification, even if only a few samples per writer are available. We use these reciprocal relationships in two ways: encode them into new vectors, as originally proposed, or integrate them in terms of query-expansion. We show that both techniques outperform the baseline results in terms of mAP on three writer identification datasets.

We extend our previous work on Inductive Conformal Prediction (ICP) for multi-label text classification and present a novel approach for addressing the computational inefficiency of the Label Powerset (LP) ICP, arrising when dealing with a high number of unique labels. We present experimental results using the original and the proposed efficient LP-ICP on two English and one Czech language data-sets. Specifically, we apply the LP-ICP on three deep Artificial Neural Network (ANN) classifiers of two types: one based on contextualised (bert) and two on non-contextualised (word2vec) word-embeddings. In the LP-ICP setting we assign nonconformity scores to label-sets from which the corresponding p -values and prediction-sets are determined. Our approach deals with the increased computational burden of LP by eliminating from consideration a significant number of label-sets that will surely have p -values below the specified significance level. This reduces dramatically the computational complexity of the approach while fully respecting the standard CP guarantees. Our experimental results show that the contextualised-based classifier surpasses the non-contextualised-based ones and obtains state-of-the-art performance for all data-sets examined. The good performance of the underlying classifiers is carried on to their ICP counterparts without any significant accuracy loss, but with the added benefits of ICP, i.e. the confidence information encapsulated in the prediction sets. We experimentally demonstrate that the resulting prediction sets can be tight enough to be practically useful even though the set of all possible label-sets contains more than 1 e + 16 combinations. Additionally, the empirical error rates of the obtained prediction-sets confirm that our outputs are well-calibrated.

This paper presents an overview of training strategies for optical character recognition of historical documents. The main issue is the lack of the annotated data and its quality. We summarize several ways of synthetic data preparation. The main goal of this paper is to show and compare possibilities how to train a convolutional recurrent neural network classifier using the synthetic data and its combination with a real annotated dataset.

This work aims at data preparation for OCR systems based on recurrent neural networks. Precisely annotated data are necessary for training a network as well as for evaluation of OCR methods. It is possible to synthesize the data, however such data are not that realistic as the real ones. Manual annotation is thus still needed in many cases, especially in the case of historical documents we are focusing on. Although there are several complex systems for historical document processing, to the best of our knowledge, a simple annotation tool for OCR data is completely missing. Therefore, we propose and implement a set of tools utilizing artificial intelligence that simplify the annotation process. These tools create ground truths for line images that are used for training of nowadays OCR systems. Another contribution of this paper is making these tools freely available for research purposes.

Článek se zabývá vícejazyčným rozpoznávání dialogových aktů. Navržené metody jsou založeny na hlubokých neuronových sítích a používají word2vec slovní vektory pro reprezentaci slov. Prezentovány jsou dvě metody. První z nich používá jeden obecný model trénovaný na všech dostupných jazycích, zatímco druhý se natrénuje pouze na jednom jazyku a pomocí lineární transformace dochází k projekci prostorů na zvolený jazyk. Jako klasifikátor jsou použity populární konvoluční neuronové sítě a LSTM. Dle našeho mínění jde o jednu z prvních metod pro vícejazyčnou klasifikaci dialogových aktů pomocí neuronových sítí. Modely jsou testovány experimentálně na dvoujazyčném korpusu Verbmobil This paper deals with multi-lingual dialogue act (DA) recognition. The proposed approaches are based on deep neural networks and use word2vec embeddings for word representation. Two multi-lingual models are proposed for this task. The first approach uses one general model trained on the embeddings from all available languages. The second method trains the model on a single pivot language and a linear transformation method is used to project other languages onto the pivot language. The popular convolutional neural network and LSTM architectures with different set-ups are used as classifiers. To the best of our knowledge this is the first attempt at multi-lingual DA recognition using neural networks. The multi-lingual models are validated experimentally on two languages from the Verbmobil corpus.

With the author(s)’ decision to order Open Choice, the copyright of the article changed on 3rd December 2020 to [The Authors] [2020] and the article is forthwith distributed under the terms of copyright.

Cross-lingual document representation can be done by training monolingual semantic spaces and then to use bilingual dictionaries with some transform method to project word vectors into a unified space. The main goal of this paper consists in evaluation of three promising transform methods on cross-lingual document classification task. We also propose, evaluate and compare two cross-lingual document classification approaches. We use popular convolutional neural network (CNN) and compare its performance with a standard maximum entropy classifier. The proposed methods are evaluated on four languages, namely English, German, Spanish and Italian from the Reuters corpus. We demonstrate that the results of all transformation methods are close to each other, however the orthogonal transformation gives generally slightly better results when CNN with trained embeddings is used. The experimental results also show that convolutional network achieves better results than maximum entropy classifier. We further show that the proposed methods are competitive with the state of the art.

This paper describes our system created for the WASSA 2018 Implicit Emotion Shared Task. The goal of this task is to predict the emotion of a given tweet, from which a certain emotion word is removed. The removed word can be sad, happy, disgusted, angry, afraid or a synonym of one of them. Our proposed system is based on deep-learning methods. We use Bidirectional Long Short-Term Memory (BiLSTM) with word embeddings as an input. Pre-trained DeepMoji model and pre-trained emoji2vec emoji embeddings are also used as additional inputs. Our System achieves 0.657 macro F1 score and our rank is 13th out of 30.

This paper proposes a novel approach for multi-lingual multi-label document classification based on neural networks. We use popular convolutional neural networks for this task with three different configurations. The first one uses static word2vec embeddings that are let as is, while the second one initializes it with word2vec and fine-tunes the embeddings while learning on the available data. The last method initializes embeddings randomly and then they are optimized to the classification task. The proposed method is evaluated on four languages, namely English, German, Spanish and Italian from the Reuters corpus. Experimental results show that the proposed approach is efficient and the best obtained F-measure reaches 84%.

The article informs readers of the monothematic issue of Geografie journal devoted to the 150th anniversary of geography at Charles University in Prague about historical roots of geography at Prague University in the context of its development in Czechia during the last 150 years. The aim of the contribution is not to assess either the history of geographical thinking or the latest history of geography, but to present above all the development of personal and institutional backing of scientific and pedagogical activities in Czech geography. When assessing this development, the authors take account of the fact that Czech geography, as a discipline of science at Charles University, was developing in relatively frequently changing and sensibly different constitutional and political systems, socio-economical and socio-cultural conditions and also in different geopolitical situations and links of the Czech state. Besides external conditions that have been determining the changes of geography, the authors stress also the role of internal (subjective) factors - existence of key personalities of the discipline (fathers founders), their capacity to get recognition in the international context and to form their continuators.