Alejandro Hector Toselli
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Alejandro Hector Toselli has a total of 56 co-authors. Here are the top co-authors by number of shared publications:
- Enrique Vidal
Universitat Politècnica de València
33 shared publications - Francisco Casacuberta
Universitat Politècnica de València
12 shared publications - Byron Leite Dantas Bezerra
Universidade de Pernambuco
7 shared publications Enrique Vidal
6 shared publications- Arthur Flor de Sousa Neto
Universidade de Pernambuco, Recife, PE, Brazil
5 shared publications - José Andrés
Universitat Politècnica de València
4 shared publications
Latest Publications
- 28 August 2026
- Published by Springer Nature
- p. 282-298
Modeling “Musical Words” for Efficient Sequence Search in Handwritten Music Manuscripts
- 28 August 2026
- Published by Springer Nature
- p. 282-298
Modeling “Musical Words” for Efficient Sequence Search in Handwritten Music Manuscripts
- Open Access
- 11 August 2026
- Published by Springer Nature in Pattern Analysis and Applications
- Vol. 29 (3) , 164
Plainchant music modeling through Zipf’s law
As important as being able to know the words in a text, being able to identify melodic patterns in musical scores is of utmost importance in many applications, such as indexing for information search and retrieval. We explore here methods to set up appropriate musical units that extend beyond individual notes into higher-level units of communication; i.e., melodic patterns, dubbed “musical words”. This prospect is studied following two main ideas for musical encoding, understood as the process of transforming musical content into “musical text” based on a musical word vocabulary: (1) based on the association of notes of musical scores and corresponding lyrics text (available in some datasets annotated with such explicit alignments); (2) based on the unsupervised grouping of individual notes into note sequences using a technique known as byte pair encoding (BPE). To assess the appropriateness of these musical encodings, we resort to Zipf’s curves of the encoded musical texts and measure how close such curves are to a Zipfian law. For the most promising vocabularies according to these tests, optical music recognition experiments are also performed. Results suggest that the musical words derived from the alignment with lyrics words produce the most natural melodic patterns and that better results are achieved if the encoding is based on note pitch intervals rather than notes encoded as absolute pitch symbols. Additionally, BPE appears to be a promising unsupervised way to encode musical content into sequences of fully automatically discovered musical words.
- 11 August 2026
- Published by Springer Nature in Pattern Analysis and Applications
- Vol. 29 (3) , 164
Open AccessPlainchant music modeling through Zipf’s law
As important as being able to know the words in a text, being able to identify melodic patterns in musical scores is of utmost importance in many applications, such as indexing for information search and retrieval. We explore here methods to set up appropriate musical units that extend beyond individual notes into higher-level units of communication; i.e., melodic patterns, dubbed “musical words”. This prospect is studied following two main ideas for musical encoding, understood as the process of transforming musical content into “musical text” based on a musical word vocabulary: (1) based on the association of notes of musical scores and corresponding lyrics text (available in some datasets annotated with such explicit alignments); (2) based on the unsupervised grouping of individual notes into note sequences using a technique known as byte pair encoding (BPE). To assess the appropriateness of these musical encodings, we resort to Zipf’s curves of the encoded musical texts and measure how close such curves are to a Zipfian law. For the most promising vocabularies according to these tests, optical music recognition experiments are also performed. Results suggest that the musical words derived from the alignment with lyrics words produce the most natural melodic patterns and that better results are achieved if the encoding is based on note pitch intervals rather than notes encoded as absolute pitch symbols. Additionally, BPE appears to be a promising unsupervised way to encode musical content into sequences of fully automatically discovered musical words.
- Open Access
- 15 July 2026
- Published by Springer Nature in Pattern Analysis and Applications
- Vol. 29 (3)
Simple handwritten text recognition techniques for highly accurate writer identification
The theater of the Spanish Early Modern period encompasses thousands of textual works and hundreds of playwrights and is one of the greatest examples of Spanish literature. These works were usually copied and altered, changing sometimes the meaning of the original works written by the authors. Therefore, identifying autograph testimonies written directly by the playwrights themselves are of particular importance. In this paper, we address an approach for writer identification based on Deep Convolutional-Recurrent Neural Networks and n -gram language models. The identification task is posed as a classification problem, introducing a probabilistic framework that goes beyond the plain transcription of handwritten text. Experiments are conducted to validate our proposal to distinguish between Lope de Vega ’s manuscripts and other non-Lope hands. The good results achieved will ultimately allow to provide modern researchers with a useful tool for cultural heritage recovery.
- 15 July 2026
- Published by Springer Nature in Pattern Analysis and Applications
- Vol. 29 (3)
Open AccessSimple handwritten text recognition techniques for highly accurate writer identification
The theater of the Spanish Early Modern period encompasses thousands of textual works and hundreds of playwrights and is one of the greatest examples of Spanish literature. These works were usually copied and altered, changing sometimes the meaning of the original works written by the authors. Therefore, identifying autograph testimonies written directly by the playwrights themselves are of particular importance. In this paper, we address an approach for writer identification based on Deep Convolutional-Recurrent Neural Networks and n -gram language models. The identification task is posed as a classification problem, introducing a probabilistic framework that goes beyond the plain transcription of handwritten text. Experiments are conducted to validate our proposal to distinguish between Lope de Vega ’s manuscripts and other non-Lope hands. The good results achieved will ultimately allow to provide modern researchers with a useful tool for cultural heritage recovery.
- Open Access
- 25 April 2026
- Published by Springer Nature in International Journal on Document Analysis and Recognition (IJDAR)
- Vol. 29 (3) , 923-939
Predicting text recognition word error rate of image documents without ground truth transcripts
Unsupervised error estimation is examined under the framework of the statistical decision theory. The error probability of a classifier trained with a training set of a finite, fixed size is considered, along with an unsupervised naive estimate of this error. In general, the naive estimates tend to be significantly smaller than the empirical errors measured using ground truth class labels, but the estimated values can be easily calibrated with a function whose parameters can be trained using moderate amounts of class-labelled data. This way, true error rates can be accurately predicted for new unlabelled data. These ideas are applied to the essential classification problem which underlies the automatic transcription of text images. As a result, various methods are developed to predict the Word Error Rate (WER) of a text image recognizer, on unseen sets of images for which no ground truth transcripts are available. Experiments on three large handwritten text datasets show that the error rates predicted by some of these methods are sufficiently accurate for practical applications. More specifically, absolute deviations of the predicted word error percentages from the corresponding real WER are lower than 1.7% for the three datasets, and the relative values of these deviations with respect to the WER of each dataset, are 5.6%, 4.2% and 6.1%.
- 25 April 2026
- Published by Springer Nature in International Journal on Document Analysis and Recognition (IJDAR)
- Vol. 29 (3) , 923-939
Open AccessPredicting text recognition word error rate of image documents without ground truth transcripts
Unsupervised error estimation is examined under the framework of the statistical decision theory. The error probability of a classifier trained with a training set of a finite, fixed size is considered, along with an unsupervised naive estimate of this error. In general, the naive estimates tend to be significantly smaller than the empirical errors measured using ground truth class labels, but the estimated values can be easily calibrated with a function whose parameters can be trained using moderate amounts of class-labelled data. This way, true error rates can be accurately predicted for new unlabelled data. These ideas are applied to the essential classification problem which underlies the automatic transcription of text images. As a result, various methods are developed to predict the Word Error Rate (WER) of a text image recognizer, on unseen sets of images for which no ground truth transcripts are available. Experiments on three large handwritten text datasets show that the error rates predicted by some of these methods are sufficiently accurate for practical applications. More specifically, absolute deviations of the predicted word error percentages from the corresponding real WER are lower than 1.7% for the three datasets, and the relative values of these deviations with respect to the WER of each dataset, are 5.6%, 4.2% and 6.1%.
- Open Access
- 7 April 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 14, 54395-54413
HTSR-Pollen: Handwritten Text Synthesis and Recognition System to Overcome Data Scarcity
Offline Handwritten Text Recognition (HTR) systems recognize and transcribe handwritten text from scanned images into digital formats. The field has become important due to the need for document digitization and data entry automation in various industries. Accurate recognition requires large and varied datasets for training optical models, where collecting and labeling these datasets is often time-consuming and impractical. To address this challenge, data augmentation and transfer learning are commonly used. Nevertheless, these traditional methods may lead to overfitting and performance degradation when data are scarce. This work proposes integrating Conditional Generative Adversarial Networks (CGANs) for data synthesis into optical model training to improve handwriting recognition in data-scarce scenarios. To validate our proposal, we conducted a study that included: (i) an exploration to establish an optimal configuration for traditional data augmentation; and (ii) extensive experiments using seven datasets. In addition, these datasets were partitioned into training subsets to simulate diverse data-scarcity conditions. Averaged over all subsets and optical models, data synthesis achieved the highest reductions compared to the baseline trained from scratch without augmentation. It reduced the Character Error Rate (CER) by 41.1% and the Word Error Rate (WER) by 28.1%. Transfer learning achieved reductions of 34.4% in CER and 23.5% in WER. Lastly, traditional data augmentation achieved reductions of 13.8% in CER and 12.3% in WER. These findings highlight the importance of data synthesis for improving HTR systems, particularly in data-scarcity contexts.
- 7 April 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 14, 54395-54413
Open AccessHTSR-Pollen: Handwritten Text Synthesis and Recognition System to Overcome Data Scarcity
Offline Handwritten Text Recognition (HTR) systems recognize and transcribe handwritten text from scanned images into digital formats. The field has become important due to the need for document digitization and data entry automation in various industries. Accurate recognition requires large and varied datasets for training optical models, where collecting and labeling these datasets is often time-consuming and impractical. To address this challenge, data augmentation and transfer learning are commonly used. Nevertheless, these traditional methods may lead to overfitting and performance degradation when data are scarce. This work proposes integrating Conditional Generative Adversarial Networks (CGANs) for data synthesis into optical model training to improve handwriting recognition in data-scarce scenarios. To validate our proposal, we conducted a study that included: (i) an exploration to establish an optimal configuration for traditional data augmentation; and (ii) extensive experiments using seven datasets. In addition, these datasets were partitioned into training subsets to simulate diverse data-scarcity conditions. Averaged over all subsets and optical models, data synthesis achieved the highest reductions compared to the baseline trained from scratch without augmentation. It reduced the Character Error Rate (CER) by 41.1% and the Word Error Rate (WER) by 28.1%. Transfer learning achieved reductions of 34.4% in CER and 23.5% in WER. Lastly, traditional data augmentation achieved reductions of 13.8% in CER and 12.3% in WER. These findings highlight the importance of data synthesis for improving HTR systems, particularly in data-scarcity contexts.
- 30 July 2025
- Published by Springer Nature
- p. 254-267
Writer Identification Using Simplified Handwritten Text Recognition Models
- 30 July 2025
- Published by Springer Nature
- p. 254-267
Writer Identification Using Simplified Handwritten Text Recognition Models
- Open Access
- 28 May 2025
- Published by Springer Nature in International Journal on Document Analysis and Recognition (IJDAR)
- Vol. 28 (3) , 377-389
The PARES Database: Information Extraction over Historical Parish Records
Historical census records convey information that is key to perform genealogical research and demographic studies. Given the large number of documents of this type that exist, it is crucial to research methods that allow the automatic extraction of information from this type of document. In this work, we present a new corpus of this kind, comprising 535 historical census tables from French archives. Alongside this dataset, we have assessed three different baseline methods for information extraction. The first two methods employ a traditional sequential approach, where table rows are detected before extracting information. The third baseline uses an end-to-end model that directly extracts information from the table images without prior row detection. Our results demonstrate the effectiveness of all three baselines in tackling the information extraction task.
- 28 May 2025
- Published by Springer Nature in International Journal on Document Analysis and Recognition (IJDAR)
- Vol. 28 (3) , 377-389
Open AccessThe PARES Database: Information Extraction over Historical Parish Records
Historical census records convey information that is key to perform genealogical research and demographic studies. Given the large number of documents of this type that exist, it is crucial to research methods that allow the automatic extraction of information from this type of document. In this work, we present a new corpus of this kind, comprising 535 historical census tables from French archives. Alongside this dataset, we have assessed three different baseline methods for information extraction. The first two methods employ a traditional sequential approach, where table rows are detected before extracting information. The third baseline uses an end-to-end model that directly extracts information from the table images without prior row detection. Our results demonstrate the effectiveness of all three baselines in tackling the information extraction task.
- 11 September 2024
- Published by Springer Nature
- p. 345-362
ICDAR 2024 Competition on Handwritten Text Recognition in Brazilian Essays – BRESSAY
- 11 September 2024
- Published by Springer Nature
- p. 345-362
ICDAR 2024 Competition on Handwritten Text Recognition in Brazilian Essays – BRESSAY
- 9 September 2024
- Published by Springer Nature
- p. 319-333
Mining and Analyzing Statistical Information from Untranscribed Form Images
- 9 September 2024
- Published by Springer Nature
- p. 319-333
Mining and Analyzing Statistical Information from Untranscribed Form Images