Search

Handwritten Text Recognition (HTR) has become an essential field within pattern recognition and machine learning, with applications spanning historical document preservation to modern data entry and accessibility solutions. The complexity of HTR lies in the high variability of handwriting, which makes it challenging to develop robust recognition systems. This survey examines the evolution of HTR models, tracing their progression from early heuristic-based approaches to contemporary state-of-the-art neural models, which leverage deep learning techniques. The scope of the field has also expanded, with models initially capable of recognizing only word-level content progressing to recent end-to-end document-level approaches. Our paper categorizes existing work into two primary levels of recognition: (1) \emph{up to line-level}, encompassing word and line recognition, and (2) \emph{beyond line-level}, addressing paragraph- and document-level challenges. We provide a unified framework that examines research methodologies, recent advances in benchmarking, key datasets in the field, and a discussion of the results reported in the literature. Finally, we identify pressing research challenges and outline promising future directions, aiming to equip researchers and practitioners with a roadmap for advancing the field.

The avoidance of water freezing is the holy grail in the cryopreservation of biological samples, food, and organs. Fast cooling rates are used to beat ice nucleation and avoid cell damage. This strategy can be enhanced by applying high pressures to decrease the nucleation rate, but the physics behind this procedure has not been fully understood yet. We perform computer experiments to investigate ice nucleation at high pressures consisting in embedding ice seeds in supercooled water. We find that the slowing down of the nucleation rate is mainly due to an increase of the ice I-water interfacial free energy with pressure. Our work also clarifies the molecular mechanism of ice nucleation for a wide pressure range. This study is not only relevant to cryopreservation, but also to water amorphization and climate change modeling., Comment: 6 pages and 3 figures

The 3D spectrophotometry measurements of the galaxy NGC~3521, a structural Milky Way analogue (sMWA), were carried out within the Metal-THINGS project. We found that the oxygen abundance in the inner part of NGC~3521 is at a nearly constant level and the O/H gradient is negative at larger radii. The change in the N/H with radius is similar to that for O/H. The radial distributions of the O/H, the gas mass fraction, and the effective oxygen yield in NGC~3521 are compared to that of the Milky Way (MW), with the aim of examining the similarity in their chemical evolutions. The O/H of two HII regions closest to the centre of the MW (at a radii of 4-5 kpc) are close to the binned O/H in NGC~3521 at the same galactocentric distances. The O/H in the outer part of the MW are lower than those in the outer part of NGC~3521. The gas mass fraction in the outer part of the MW is higher than in NGC~3521. The obtained values of the effective oxygen yield, Yeff, in NGC~3521 are close to the empirical estimation of the oxygen yield, Yo. This suggests that mass exchange with the surroundings plays little to no role in the chemical evolution of NGC3521. The values of the Yeff in the MW were determined using two variants of the distribution of the gas mass surface density. The values of the Yeff in the MW obtained with the first distribution are also close to Yo. The Yeff in the MW obtained with the second distribution are below Yo at radii between 6 and 10.4 kpc. This suggests that the mass exchange with the surroundings can play a significant role in the chemical evolution of this part of the MW. To draw a solid conclusion about the role of mass exchange with the surroundings in the chemical evolution of the MW it is essential to determine which of these distributions provides a more adequate description of the gas distribution in the MW., Comment: 15 pages, 12 figures, accepted for publication in Astron.Astrophys

In recent years, polar codes have been considered for communication systems that require high re-liability and ultra-low latency, such as sixth generation (6G) wireless communications. This paper presents simulation results showing that short binary extended BCH (eBCH) codes with low-complexity decoding outperform polar codes for lengths 64 and 128. In the simulations, polar mapping under additive white Gaussian noise (AWGN) is assumed and ordered-statistics decoding (OSD) of eBCH codes is compared with CRC-aided successive-cancellation list decoding (SCLD-CRC) of polar codes of the same lengths and rates. The results indicate that short-length binary eBCH codes achieve lower average bit error rate values (higher reliability) and thus should be considered as strong candidates in communication systems requiring extremely low latency, i.e., short code lengths of up to 128 bits. The eBCH simulation results are obtained with OSD and re-processing order equal to one so that complexity is comparable to SCLD-CRC. Specifically, error performances are quantified of length-64 and selected length-128 eBCH codes with order-1 OSD and polar codes with SCLD-CRC for the same rates and lengths. These results serve to verify that short binary eBCH codes do indeed outperform short polar codes with comparable decoding complexity.

The digitization of vocal music scores presents unique challenges that go beyond traditional Optical Music Recognition (OMR) and Optical Character Recognition (OCR), as it necessitates preserving the critical alignment between music notation and lyrics. This alignment is essential for proper interpretation and processing in practical applications. This paper introduces and formalizes, for the first time, the Aligned Music Notation and Lyrics Transcription (AMNLT) challenge, which addresses the complete transcription of vocal scores by jointly considering music symbols, lyrics, and their synchronization. We analyze different approaches to address this challenge, ranging from traditional divide-and-conquer methods that handle music and lyrics separately, to novel end-to-end solutions including direct transcription, unfolding mechanisms, and language modeling. To evaluate these methods, we introduce four datasets of Gregorian chants, comprising both real and synthetic sources, along with custom metrics specifically designed to assess both transcription and alignment accuracy. Our experimental results demonstrate that end-to-end approaches generally outperform heuristic methods in the alignment challenge, with language models showing particular promise in scenarios where sufficient training data is available. This work establishes the first comprehensive framework for AMNLT, providing both theoretical foundations and practical solutions for preserving and digitizing vocal music heritage.

Recent advances in Handwritten Text Recognition (HTR) have led to significant reductions in transcription errors on standard benchmarks under the i.i.d. assumption, thus focusing on minimizing in-distribution (ID) errors. However, this assumption does not hold in real-world applications, which has motivated HTR research to explore Transfer Learning and Domain Adaptation techniques. In this work, we investigate the unaddressed limitations of HTR models in generalizing to out-of-distribution (OOD) data. We adopt the challenging setting of Domain Generalization, where models are expected to generalize to OOD data without any prior access. To this end, we analyze 336 OOD cases from eight state-of-the-art HTR models across seven widely used datasets, spanning five languages. Additionally, we study how HTR models leverage synthetic data to generalize. We reveal that the most significant factor for generalization lies in the textual divergence between domains, followed by visual divergence. We demonstrate that the error of HTR models in OOD scenarios can be reliably estimated, with discrepancies falling below 10 points in 70\% of cases. We identify the underlying limitations of HTR models, laying the foundation for future research to address this challenge.

The International Workshop on Reading Music Systems (WoRMS) is a workshop that tries to connect researchers who develop systems for reading music, such as in the field of Optical Music Recognition, with other researchers and practitioners that could benefit from such systems, like librarians or musicologists. The relevant topics of interest for the workshop include, but are not limited to: Music reading systems; Optical music recognition; Datasets and performance evaluation; Image processing on music scores; Writer identification; Authoring, editing, storing and presentation systems for music scores; Multi-modal systems; Novel input-methods for music to produce written music; Web-based Music Information Retrieval services; Applications and projects; Use-cases related to written music. These are the proceedings of the 6th International Workshop on Reading Music Systems, held Online on November 22nd 2024., Comment: Proceedings edited by Jorge Calvo-Zaragoza, Alexander Pacha and Elona Shatri

The speed of response by search and rescue teams at sea is of vital importance, as survival may depend on it. Recent technological advancements have led to the development of more efficient systems for locating individuals involved in a maritime incident, such as the use of Unmanned Aerial Vehicles (UAVs) equipped with cameras and other integrated sensors. Over the past decade, several researchers have contributed to the development of automatic systems capable of detecting people using aerial images, particularly by leveraging the advantages of deep learning. In this article, we provide a comprehensive review of the existing literature on this topic. We analyze the methods proposed to date, including both traditional techniques and more advanced approaches based on machine learning and neural networks. Additionally, we take into account the use of synthetic data to cover a wider range of scenarios without the need to deploy a team to collect data, which is one of the major obstacles for these systems. Overall, this paper situates the reader in the field of detecting people at sea using aerial images by quickly identifying the most suitable methodology for each scenario, as well as providing an in-depth discussion and direction for future trends.

Text Recognition (TR) refers to the research area that focuses on retrieving textual information from images, a topic that has seen significant advancements in the last decade due to the use of Deep Neural Networks (DNN). However, these solutions often necessitate vast amounts of manually labeled or synthetic data. Addressing this challenge, Self-Supervised Learning (SSL) has gained attention by utilizing large datasets of unlabeled data to train DNN, thereby generating meaningful and robust representations. Although SSL was initially overlooked in TR because of its unique characteristics, recent years have witnessed a surge in the development of SSL methods specifically for this field. This rapid development, however, has led to many methods being explored independently, without taking previous efforts in methodology or comparison into account, thereby hindering progress in the field of research. This paper, therefore, seeks to consolidate the use of SSL in the field of TR, offering a critical and comprehensive overview of the current state of the art. We will review and analyze the existing methods, compare their results, and highlight inconsistencies in the current literature. This thorough analysis aims to provide general insights into the field, propose standardizations, identify new research directions, and foster its proper development., Comment: This article is under revision

The tilt-to-length coupling during the LISA Pathfinder mission has been numerically and analytically modeled for particular timespans. In this work, we investigate the long-term stability of the coupling coefficients of this noise. We show that they drifted slowly (by 1\,$\mu$m/rad and 6$\times10^{-6}$ in 100 days) and strongly correlated to temperature changes within the satellite (8\,$\mu$m/rad/K and 30$\times10^{-6}$/K). Based on analytical TTL coupling models, we attribute the temperature-driven coupling changes to rotations of the test masses and small distortions in the optical setup. Particularly, we show that LISA Pathfinder's optical baseplate was bent during the cooldown experiment, which started in late 2016 and lasted several months.

Glycerol acts as a natural cryoprotectant by depressing the temperature of ice nucleation and slowing down the dynamics of water mixtures. In this work we characterize dynamics -- diffusion, viscosity, and hydrogen-bond dynamics -- as well as density anomaly and structure of water mixtures with 1\% to 50\% w/w glycerol at low temperatures via molecular dynamics simulations using all-atom and coarse-grained models. Simulations reveal distinct violations of the Stokes-Einsten relation in the low temperature regime for water and glycerol. Deviations are positive for water at all concentrations, and positive for glycerol in very dilute solutions but turning negative in concentrated ones. The all-atom and coarse-grained models reveal an unexpected crossover in the dynamics of the 1% and 10 % w/w glycerol at the lowest simulated temperatures. This crossover manifests in the diffusion coefficients of water and glycerol, as well as in the viscosity and lifetime of hydrogen-bonds in water. We interpret that the crossover originates on the opposing dependence with glycerol concentration of the two factors controlling the solution's slow-down: the increase in tetrahedrally coordinated water and the dynamics and clustering of the glycerol molecules. We anticipate that this dynamic crossover will also occur for solution of water with other polyols., Comment: Anomalies; Transport properties; Mixtures; Water; Viscosity; Simulations

A precise characterization of the magnetic properties of LISA Pathfinder free falling test-masses is of special interest for future gravitational wave observatory in space. Magnetic forces have an important impact on the instrument sensitivity in the low frequency regime below the millihertz. In this paper we report on the magnetic injection experiments performed throughout LISA Pathfinder operations. We show how these experiments allowed a high precision estimate of the instrument magnetic parameters. The remanent magnetic moment was found to have a modulus of $(0.245\pm0.081)\,\rm{nAm}^2$, the x-component of the background magnetic field within the test masses position was measured to be $(414 \pm 74)$ nT and its gradient had a value of $(-7.4\pm 2.1)\,\mu$T/m. Finally, we also measured the test mass magnetic susceptibility to be $(-3.35\pm0.15)\times$10$^{-5}$ in the low frequency regime. All results are in agreement with on-ground estimates.

LISA Pathfinder was a mission designed to test key technologies required for gravitational wave detection in space. Magnetically driven forces play a key role in the instrument sensitivity in the low-frequency regime, which corresponds to the measurement band of interest for future space-borne gravitational wave observatories. Magnetic-induced forces couple to the test mass motion, introducing a contribution to the relative acceleration noise between the free falling test masses. In this Letter we present the first complete estimate of this term of the instrument performance model. Our results set the magnetic-induced acceleration noise during the February 2017 noise run of $\rm 0.25_{-0.08}^{+0.15}\,fm\,s^{-2}/\sqrt{Hz}$ at 1 mHz and $\rm 1.01_{-0.24}^{+0.73}\, fm\,s^{-2}/\sqrt{Hz}$ at 0.1 mHz. We also discuss how the non-stationarities of the interplanetary magnetic field can affect these values during extreme space weather conditions.

Tidal interactions between galaxies often give rise to tidal tails, which can harbor concentrations of stars and interstellar gas resembling dwarf galaxies. Some of these tidal dwarf galaxies (TDGs) have the potential to detach from their parent galaxies and become independent entities, but their long-term survival is uncertain. In this study, we conducted a search for detached TDGs associated with a sample of 39 interacting galaxy pairs in the local Universe using infrared, ultraviolet, and optical images. We employed IR colors and UV/optical/IR spectral energy distributions to identify potential interlopers, such as foreground stars or background quasars. Through spectroscopic observations using the Boller and Chivens spectrograph at San Pedro M\'artir Observatory, we confirmed that six candidate TDGs are at the same redshift as their putative parent galaxy pairs. We identified and measured emission lines in the optical spectra and calculated nebular oxygen abundances, which range from log(O/H) = 8.10 $\pm$ 0.01 to 8.51 $\pm$ 0.02. We have serendipitously discovered an additional detached TDG candidate in Arp72 using available spectra from SDSS. Utilizing the photometric data and the CIGALE code for stellar population and dust emission fitting, we derived the stellar masses, stellar population ages, and stellar metallicities for these detached TDGs. Compared to standard mass-metallicity relations for dwarf galaxies, five of the seven candidates have higher than expected metallicities, confirming their tidal origins. One of the seven candidates remains unclear due to large uncertainties in metallicity, and another has stellar and nebular metallicities compatible with those of a preexisting dwarf galaxy. The latter object is relatively compact in the optical relative to its stellar mass, in contrast to the other candidate TDGs [abridged]., Comment: Accepted for publication in Astronomy and Astrophysics. 15 pages, 9 figures, 6 tables

Optical Music Recognition (OMR) has made significant progress since its inception, with various approaches now capable of accurately transcribing music scores into digital formats. Despite these advancements, most so-called \emph{end-to-end} OMR approaches still rely on multi-stage processing pipelines for transcribing full-page score images, which introduces several limitations that hinder the full potential of the field. In this paper, we present the first truly end-to-end approach for page-level OMR. Our system, which combines convolutional layers with autoregressive Transformers, processes an entire music score page and outputs a complete transcription in a music encoding format. This is made possible by both the architecture and the training procedure, which utilizes curriculum learning through incremental synthetic data generation. We evaluate the proposed system using pianoform corpora. This evaluation is conducted first in a controlled scenario with synthetic data, and subsequently against two real-world corpora of varying conditions. Our approach is compared with leading commercial OMR software. The results demonstrate that our system not only successfully transcribes full-page music scores but also outperforms the commercial tool in both zero-shot settings and after fine-tuning with the target domain, representing a significant contribution to the field of OMR.

We present an in-depth analysis of the LISA Pathfinder differential acceleration performance over the entire course of its science operations, spanning approximately 500 days. We find that: 1) the evolution of the Brownian noise that dominates the acceleration amplitude spectral density (ASD), for frequencies $f\gtrsim 1\,\text{mHz}$, is consistent with the decaying pressure due to the outgassing of a single gaseous species. 2) between $f=36\,\mu\text{Hz}$ and $1\,\text{mHz}$, the acceleration ASD shows a $1/f$ tail in excess of the Brownian noise of almost constant amplitude, with $\simeq 20\%$ fluctuations over a period of a few days, with no particular time pattern over the course of the mission; 3) at the lowest considered frequency of $f=18\,\mu\text{Hz}$, the ASD significantly deviates from the $1/f$ behavior, because of temperature fluctuations that appear to modulate a quasi-static pressure gradient, sustained by the asymmetries of the outgassing pattern. We also present the results of a projection of the observed acceleration noise on the potential sources for which we had either a direct correlation measurement, or a quantitative estimate from dedicated experiments. These sources account for approximately $40\%$ of the noise power in the $1/f$ tail. Finally, we analyze the possible sources of the remaining unexplained fraction, and identify the possible measures that may be taken to keep those under control in LISA.

This paper serves to introduce the Align, Minimize and Diversify (AMD) method, a Source-Free Unsupervised Domain Adaptation approach for Handwritten Text Recognition (HTR). This framework decouples the adaptation process from the source data, thus not only sidestepping the resource-intensive retraining process but also making it possible to leverage the wealth of pre-trained knowledge encoded in modern Deep Learning architectures. Our method explicitly eliminates the need to revisit the source data during adaptation by incorporating three distinct regularization terms: the Align term, which reduces the feature distribution discrepancy between source and target data, ensuring the transferability of the pre-trained representation; the Minimize term, which encourages the model to make assertive predictions, pushing the outputs towards one-hot-like distributions in order to minimize prediction uncertainty, and finally, the Diversify term, which safeguards against the degeneracy in predictions by promoting varied and distinctive sequences throughout the target data, preventing informational collapse. Experimental results from several benchmarks demonstrated the effectiveness and robustness of AMD, showing it to be competitive and often outperforming DA methods in HTR., Comment: Submitted to ECCV 2024

Handwritten Text Recognition (HTR) is a relevant problem in computer vision, and implies unique challenges owing to its inherent variability and the rich contextualization required for its interpretation. Despite the success of Self-Supervised Learning (SSL) in computer vision, its application to HTR has been rather scattered, leaving key SSL methodologies unexplored. This work focuses on one of them, namely Spatial Context-based SSL. We investigate how this family of approaches can be adapted and optimized for HTR and propose new workflows that leverage the unique features of handwritten text. Our experiments demonstrate that the methods considered lead to advancements in the state-of-the-art of SSL for HTR in a number of benchmark cases.

Language identification is a crucial component in the automated production of language resources, particularly in multilingual and big data contexts. However, commonly used language identifiers struggle to differentiate between similar or closely-related languages. This paper introduces FastSpell, a language identifier that combines fastText (a pre-trained language identifier tool) and Hunspell (a spell checker) with the aim of having a refined second-opinion before deciding which language should be assigned to a text. We provide a description of the FastSpell algorithm along with an explanation on how to use and configure it. To that end, we motivate the need of such a tool and present a benchmark including some popular language identifiers evaluated during the development of FastSpell. We show how FastSpell is useful not only to improve identification of similar languages, but also to identify new ones ignored by other tools.

Large-scale face recognition datasets are collected by crawling the Internet and without individuals' consent, raising legal, ethical, and privacy concerns. With the recent advances in generative models, recently several works proposed generating synthetic face recognition datasets to mitigate concerns in web-crawled face recognition datasets. This paper presents the summary of the Synthetic Data for Face Recognition (SDFR) Competition held in conjunction with the 18th IEEE International Conference on Automatic Face and Gesture Recognition (FG 2024) and established to investigate the use of synthetic data for training face recognition models. The SDFR competition was split into two tasks, allowing participants to train face recognition systems using new synthetic datasets and/or existing ones. In the first task, the face recognition backbone was fixed and the dataset size was limited, while the second task provided almost complete freedom on the model backbone, the dataset, and the training pipeline. The submitted models were trained on existing and also new synthetic datasets and used clever methods to improve training with synthetic data. The submissions were evaluated and ranked on a diverse set of seven benchmarking datasets. The paper gives an overview of the submitted face recognition models and reports achieved performance compared to baseline models trained on real and synthetic datasets. Furthermore, the evaluation of submissions is extended to bias assessment across different demography groups. Lastly, an outlook on the current state of the research in training face recognition models using synthetic data is presented, and existing problems as well as potential future directions are also discussed., Comment: The 18th IEEE International Conference on Automatic Face and Gesture Recognition (FG 2024)

We present the HPLT (High Performance Language Technologies) language resources, a new massive multilingual dataset including both monolingual and bilingual corpora extracted from CommonCrawl and previously unused web crawls from the Internet Archive. We describe our methods for data acquisition, management and processing of large corpora, which rely on open-source software tools and high-performance computing. Our monolingual collection focuses on low- to medium-resourced languages and covers 75 languages and a total of ~5.6 trillion word tokens de-duplicated on the document level. Our English-centric parallel corpus is derived from its monolingual counterpart and covers 18 language pairs and more than 96 million aligned sentence pairs with roughly 1.4 billion English tokens. The HPLT language resources are one of the largest open text corpora ever released, providing a great resource for language modeling and machine translation training. We publicly release the corpora, the software, and the tools used in this work., Comment: LREC-COLING 2024

Spontaneous datasets for Speech Emotion Recognition (SER) are scarce and frequently derived from laboratory environments or staged scenarios, such as TV shows, limiting their application in real-world contexts. We developed and publicly released the Emotional Voice Messages (EMOVOME) dataset, including 999 voice messages from real conversations of 100 Spanish speakers on a messaging app, labeled in continuous and discrete emotions by expert and non-expert annotators. We evaluated speaker-independent SER models using acoustic features as baseline and transformer-based models. We compared the results with reference datasets including acted and elicited speech, and analyzed the influence of annotators and gender fairness. The pre-trained UniSpeech-SAT-Large model achieved the highest results, 61.64% and 55.57% Unweighted Accuracy (UA) for 3-class valence and arousal prediction respectively on EMOVOME, a 10% improvement over baseline models. For the emotion categories, 42.58% UA was obtained. EMOVOME performed lower than the acted RAVDESS dataset. The elicited IEMOCAP dataset also outperformed EMOVOME in predicting emotion categories, while similar results were obtained in valence and arousal. EMOVOME outcomes varied with annotator labels, showing better results and fairness when combining expert and non-expert annotations. This study highlights the gap between controlled and real-life scenarios, supporting further advancements in recognizing genuine emotions., Comment: This article is a merged version of the description of the EMOVOME database in arXiv:2402.17496v1 and the speech emotion recognition models in arXiv:2403.02167v1. This work has been submitted to the IEEE for possible publication