Search

Handwriting typically consists of a wide range of writing forms with substantial differences in the placements and size of those writing shapes. The arrangement, organization, and spatial association of individual letters, words, and phrases on a written page are all examples of handwritten text structure. This study proposed a novel approach in which a handwritten text structure is primarily transformed into a Cartesian XY-coordinate system while maintaining its shape characteristics. The points on the contour of handwritten text with local maximum bending values are considered hypothetical breakdown points called Key Points (KPs). From the skeletal representation of handwriting, this approach first identifies KPs and their coordinate positions. After that, KPs are transformed into a graph-based representation with vertices and edges. With the use of geometric graphs, this representation attempts to capture the spatial and temporal organization of handwriting. Previous techniques focused on statistical methods, which offer fixed-size descriptions; however, graph-based representations are flexible in size and reveal the relationship between text structure. Graph representations enable the incorporation of contextual information by integrating extra features which improve text accuracy and understanding. The proposed approach is tested for writer identification and verification tasks on four benchmark datasets (CERUG-EN, CVL, Firemaker, and IAM) and one custom-built dataset. The findings demonstrated that the proposed approach obtained state-of-the-art accuracies across all the mentioned datasets. Particularly, it achieved the highest accuracy of 99.75% for identification and 99.81% for verification on the Firemaker dataset. [ABSTRACT FROM AUTHOR]

Writer identification form historical manuscripts presents a challenging problem with significant implications for understanding the authorship of ancient texts. In this paper, we propose a novel deep learning framework tailored for the task of historical manuscripts writer identification. Our approach leverages data-driven features, harnessing the power of neural networks to extract and learn discriminative patterns from handwritten historical documents. The key innovation of our framework lies in its ability to automatically discover and utilize relevant features from data to profile the writer, eliminating the need for manual feature engineering. Our methodology encompasses three well-defined steps: initially, manuscript preprocessing involves image denoising using advanced techniques such as non-local means and total-variation, followed by binarization using a Canny-edge detector. In the subsequent phase, we employ Harris corner detector for automatic key-point detection and clustering, allowing us to identify the regions of interest within the documents. Lastly, the features extracted from these regions are subjected to classification through transfer learning, utilizing a deep learning-based model specifically trained on the extracted patches. To achieve the final document-level identification, we enhance the system accuracy by implementing a majority vote scheme, where the aggregated decisions from multiple patches contribute to the ultimate classification outcome. We validate our approach on "ICDAR 2017" dataset, spanning different periods and writing styles of historical manuscripts. Experimental results demonstrate the superior performance of our method in accurately identifying the authors of historical documents, surpassing existing techniques. Moreover, our framework exhibits robustness in scenarios where limited training data is available. This work not only contributes to the field of historical manuscripts analysis but also highlights the potential of deep learning in solving intricate problems in the realm of document analysis and authorship attribution. Our framework offers a promising avenue for scholars and historians to gain deeper insights into the authors of historical texts, opening new doors for historical research and preservation. [ABSTRACT FROM AUTHOR]

Since the dawn of civilization, handwriting has been one of the most important forms of communication. However, as handwriting differs from person to person, writer identification has become a promising application of pattern recognition to identify the actual writer of a handwritten document. Handwriting can be either online or offline, depending on how it was obtained. Users can write directly on tablets, smartphones, touch screens, PDAs, and other devices using input devices with online handwriting, whereas offline handwriting is done with a pen and paper. With the advent of artificial intelligence, and most importantly deep learning techniques, the development of writer identification systems based on offline handwritten documents has gained a lot of attention. Deep learning models have the capability of automatic feature extraction, which results in increased performance. From the literature survey, it was revealed that least attention has been paid towards the development of deep learning-based writer identification systems for offline Urdu handwritten documents, unlike the English and Arabic scripts. Therefore, in this paper, we proposed an offline Urdu handwritten writer identification system using a deep learning model inspired by the VGG-16 model of CNN. The model was trained and tested on a novel Urdu handwritten dataset contributed by 318 distinct Urdu writers, resulting in an overall training accuracy of 98.71% and a testing accuracy of 99.11%. The results achieved showed that the proposed model outperformed the already existing writer identification techniques. [ABSTRACT FROM AUTHOR]

The extraction of paleographical features is an important task to study the identity of the text in the Historical Manuscripts. One of the major features is the identification of the writer or copyist. Many researchers have worked on an automated system for writer identification, and with the development of deep learning techniques many approaches have been proposed. Most of the previous studies have developed a multi-steps system, while very few of them performed an End-to-End approach. Most of the systems rely on a pre-processing step to prepare the data in order to facilitate recognition. This paper presents an End-to-End deep learning system for writer identification, tested on four different datasets: ICDAR19 and ICFHR20 (Latin datasets), KHATT and Balamand (Arabic datasets). The system is based on the Deep-TEN approach using a customized ResNet-50 network for features and local descriptor extraction with an integration of a NetVLAD end-layer to compute and encode the global descriptor. It was compared with our state-of-the-art system, winner of ICFHR20 HisFrag competition, and showed an interesting performance on all datasets without any pre-processing techniques. [ABSTRACT FROM AUTHOR]

In this work, we present two advanced models for identifying script writers, leveraging the power of deep learning. The proposed systems utilize the new vision Swin Transformer and ResNeSt-50. Swin Transformer is known for its robustness to variations and ability to model long-range dependencies, which helps capture context and make robust predictions. Through extensive training on large datasets of handwritten text samples, the Swin Transformer operates on sequences of image patches and learns to establish a robust representation of each writer’s unique style. On the other hand, ResNeSt-50 (Residual Neural Network with Squeeze-and-Excitation (SE) and Next Stage modules), with its multiple layers, helps in learning complex representations of a writer’s unique style and distinguishing between different writing styles with high precision. The SE module within ResNeSt helps the model focus on distinctive handwriting characteristics and reduce noise. The experimental results demonstrate exceptional performance, achieving an accuracy of 98.50% (at patch level) by the Swin Transformer on the CVL database, which consists of images with cursively handwritten German and English texts, and an accuracy of 96.61% (at page level) by ResNeSt-50 on the same database. This research advances writer identification by showcasing the effectiveness of the Swin Transformer and ResNeSt-50. The achieved accuracy underscores the potential of these models to process and understand complex handwriting effectively.

In recent years, there has been a significant increase in document forgery, which includes the fraudulent replication of currency, diplomas, and works of art. This has become a major issue due to the widespread usage of paper-based documentation. Handwriting is closely linked to document forgery and forensics as it possesses unique characteristics, including variations in text characters, pen pressure, writing angle, and stroke patterns, which makes it impossible to replicate accurately. As a result, handwriting serves as a personalized biometric that can be used to determine the authenticity of documents. However, traditional methods of writer identification are both time-consuming and destructive, requiring substantial expertise. To overcome these limitations, the study explores the potential of hyperspectral imaging (HSI) as a non-destructive and advanced approach for detecting and preventing document forgery. HSI provides detailed spectral information from a scene, making it possible to capture subtle spectral differences in handwriting samples. This imaging technique has diverse applications in various fields such as agriculture, environmental monitoring, remote sensing, forensics, document analysis, and medical imaging. Our study proposes a novel unsupervised approach, CAE-SVM that uses Convolutional Autoencoder (CAE) for feature extraction and Support Vector Machine (SVM) for writer identification. It was tested on the UWA writing ink hyperspectral images dataset for blue and black inks which is available publicly and compared with state-of-the-art methods and CNN. The proposed approach achieved the highest accuracy of 92.78% for blue ink, surpassing existing methods. The study's results emphasize the efficacy of HSI as a potent forensic analysis tool for detecting and preventing document forgery. [ABSTRACT FROM AUTHOR]

Handwriting is an individual trait that serves as evidence to authenticate a particular writer. Identifying the writer of a handwritten text has shown encouraging results in examining historical and forensic documents. In this paper, we propose a novel offline writer identification system based on the challenging analysis of small amount of data to extract distinct patterns. In our deep network, the feature extraction process relies on a specially designed dual-path architecture, and the resulting embeddings are concatenated to produce the final learned features. To deal with a variety of uncertainties such as high intra-class variations and noises, we leverage the fuzzy logic in the design of a custom Convolutional Neural Network (CNN) with a type-2 fuzzy activation function for the first path. Additionally, the second path utilizes the transfer learning-based CNN to enhance the discriminability of the learned features. Our method allows for text-independent writer identification, eliminating the need for identical handwriting samples to train and test the model. Considering that various factors can influence the handwriting style, a dataset containing right-to-left handwriting samples is assembled. The proposed method is evaluated on our developed dataset and four widely-known public datasets, namely KHATT, CVL, Firemaker, and IAM. High accuracy values are achieved, with results of 99.85%, 99.83%, 99.79%, 99.64%, and 98.17% for each dataset, respectively. One noteworthy aspect of this study is that the evaluation results on diverse datasets demonstrate the applicability of the proposed model to various languages. Moreover, the model performs effectively in real-world scenarios with limited handwritten data. [ABSTRACT FROM AUTHOR]

The writer identification task infers the writer by analyzing the texture, structure, and other representative features of the handwriting. Inspired by the attention mechanism, an end-to-end writer identification model is proposed in this paper, which combines both global features and local features. The Vision Transformer is used as the backbone network, and the Convolutional block attention module (CBAM) is introduced to enhance the ability of global feature awareness of the model. The proposed method is evaluated on two public data sets, IAM and CVL respectively. In the task of word-level writer identification, the accuracy rates in two data sets were 90.1% and 92.3% respectively. In the task of page-level writer identification, the accuracy rates were 98.6% and 99.5%, as a state-of-the-art performance. [ABSTRACT FROM AUTHOR]

Signature is used as an important, typical means of identification in banking, security controls, certificates, and contracts. Moreover, it has an intensive role in legal issues that need to be studied, considering the development of related applications. From this standpoint, this study proposes a new method for identifying offline signatures under different uncertainties such as various experimental conditions and environmental noises using deep learning approaches. To achieve this objective, a comprehensive right-to-left signature dataset based on relevant standards is collected from 85 participants at various time intervals under different experimental conditions. A developed deep neural network based on transfer learning networks is designed to extract features from raw data in a hierarchical manner. One of the benefits of the proposed method is that it is independent of right-handed or left-handed people and can be applied to both. The method proposed is examined not just on the collected dataset but also on a variety of other datasets. The proposed network has a 99% accuracy for author signature classification and can withstand a wide range of SNRs. As a result, the classification accuracy at 15 dB remains above 90%. The study's findings show that the proposed network can learn features hierarchically from raw signature data and achieve greater accuracy than other methods. Because of its superior performance, the proposed model can be used as an assistant by signature experts in a variety of applications, including the detection of forgery and criminals. [ABSTRACT FROM AUTHOR]

Abstract The main goal of the present work is to determine the hand that has written two newly discovered documents in Romania. For giving the proper answer, the authors introduced the notion of “Ideal Representative”, namely of an object that very well represents the corresponding ideal alphabet symbol that a writer had in his/her mind when writing a document by hand. Moreover, the authors have introduced a novel method, which leads to the optimal evaluation of the Ideal Representative of any alphabet symbol in association with any handwritten document. Furthermore, the authors have introduced methods for comparing these Ideal Representatives, so as a final decision about the hand that has written a document may be obtained with a highly considerable likelihood. The related analysis manifests that the two documents discovered in Romania in 1998, belong to the great personality of Rigas Feraios. The presented method of automatic handwriting Identification seems to be of general applicability.

Identifying the writer of a handwritten document has remained an interesting pattern classification problem for document examiners, forensic experts, and paleographers. While mature identification systems have been developed for handwriting in contemporary documents, the problem remains challenging from the viewpoint of historical manuscripts. Design and development of expert systems that can identify the writer of a questioned manuscript or retrieve samples belonging to a given writer can greatly help the paleographers in their practices. In this context, the current study exploits the textural information in handwriting to characterize writer from historical documents. More specifically, we employ oBIF(oriented Basic Image Features) and hinge features and introduce a novel moment-based matching method to compare the feature vectors extracted from writing samples. Classification is based on minimization of a similarity criterion using the proposed moment distance. A comprehensive series of experiments using the International Conference on Document Analysis and Recognition 2017 historical writer identification dataset reported promising results and validated the ideas put forward in this study.

• We consider large-scale supervised pre-training on a carefully designed synthetic dataset of word images. • We obtain robust writer's style representations, independent of the semantic content of the image. • We leverage the obtained representations for handwriting analysis tasks on real images from benchmark datasets. • Experiments demonstrate the suitability of our approach and its competitiveness compared to task-specific state-of-the-art. In this work, we explore massive pre-training on synthetic word images for enhancing the performance on four benchmark downstream handwriting analysis tasks. To this end, we build a large synthetic dataset of word images rendered in several handwriting fonts, which offers a complete supervision signal. We use it to train a simple convolutional neural network (ConvNet) with a fully supervised objective. The vector representations of the images obtained from the pre-trained ConvNet can then be considered as encodings of the handwriting style. We exploit such representations for Writer Retrieval, Writer Identification, Writer Verification, and Writer Classification and demonstrate that our pre-training strategy allows extracting rich representations of the writers' style that enable the aforementioned tasks with competitive results with respect to task-specific State-of-the-Art approaches. [ABSTRACT FROM AUTHOR]

The writer identification system identifies individuals based on their handwriting is a frequent topic in biometric authentication and verification systems. Due to its importance, numerous studies have been conducted in various languages. Researchers have established several learning methods for writer identification including supervised and unsupervised learning. However, supervised methods require a large amount of annotation data, which is impossible in most scenarios. On the other hand, unsupervised writer identification methods may be limited and dependent on feature extraction that cannot provide the proper objectives to the architecture and be misinterpreted. This paper introduces an unsupervised writer identification system that analyzes the data and recognizes the writer based on the inter-feature relations of the data to resolve the uncertainty of the features. A pairwise architecturebased Autoembedder was applied to generate clusterable embeddings for handwritten text images. Furthermore, the trained baseline architecture generates the embedding of the data image, and the K-means algorithm is used to distinguish the embedding of individual writers. The proposed model utilized the IAM dataset for the experiment as it is inconsistent with contributions from the authors but is easily accessible for writer identification tasks. In addition, traditional evaluation metrics are used in the proposed model. Finally, the proposed model is compared with a few unsupervised models, and it outperformed the state-of-the-art deep convolutional architectures in recognizing writers based on unlabeled data. [ABSTRACT FROM AUTHOR]

As an indispensable part of identity authentication, offline writer identification plays a notable role in biology, forensics, and historical document analysis. However, identifying handwriting efficiently, stably, and quickly is still challenging due to the method of extracting and processing handwriting features. In this paper, we propose an efficient system to identify writers through handwritten images, which integrates local and global features from similar handwritten images. The local features are modeled by effective aggregate processing, and global features are extracted through transfer learning. Specifically, the proposed system employs a pre-trained Residual Network to mine the relationship between large image sets and specific handwritten images, while the vector of locally aggregated descriptors with double power normalization is employed in aggregating local and global features. Moreover, handwritten image segmentation, preprocessing, enhancement, optimization of neural network architecture, and normalization for local and global features are exploited, significantly improving system performance. The proposed system is evaluated on Computer Vision Lab (CVL) datasets and the International Conference on Document Analysis and Recognition (ICDAR) 2013 datasets. The results show that it represents good generalizability and achieves state-of-the-art performance. Furthermore, the system performs better when training complete handwriting patches with the normalization method. The experimental result indicates that it’s significant to segment handwriting reasonably while dealing with handwriting overlap, which reduces visual burstiness. [ABSTRACT FROM AUTHOR]

Offline handwriting has become a very important part in the field of behavioral biometrics. In this paper, we proposed an approach to identify writers from their handwritten documents. The foremost contribution of this study is to suggest the use of co-occurrence features to improve the performance of writer identification. A contour texture-based feature is extracted from preprocessed regions of interest (components or sub-images) and their exterior contours are based primarily on Modified Local Binary Pattern (MLBP) and Ink-trace Width and Shape Letters (IWSL) measurements. Considering the contour as a texture, and using these textural descriptors, the joint probability distributions of MLBP and IWSL on different pixels are calculated in order to determine the similarities between different images of handwriting. Identification is carried out using the nearest neighbor rule and Chi-square distance. The proposed system has been evaluated on eight well-known handwriting databases (Arabic IFN/ENIT and KHATT, English IAM and CVL, Dutch Firemaker, Portuguese BFL, Chinese CERUG-CN, and English/Greek ICDAR2013). Experimental results show that the recommended scheme achieves the highest performance on KHATT, CVL, Firemaker, BFL, CERUG-CN, and ICDAR2013 databases, and that it demonstrates competitive performance on IFN/ENIT and IAM databases as compared to those reported by the state-of-the-art identification systems.

The main goal of the present work is to determine the hand that has written two newly discovered documents in Romania. For giving the proper answer, the authors introduced the notion of "Ideal Representative", namely of an object that very well represents the corresponding ideal alphabet symbol that a writer had in his/her mind when writing a document by hand. Moreover, the authors have introduced a novel method, which leads to the optimal evaluation of the Ideal Representative of any alphabet symbol in association with any handwritten document. Furthermore, the authors have introduced methods for comparing these Ideal Representatives, so as a final decision about the hand that has written a document may be obtained with a highly considerable likelihood. The related analysis manifests that the two documents discovered in Romania in 1998, belong to the great personality of Rigas Feraios. The presented method of automatic handwriting Identification seems to be of general applicability. [ABSTRACT FROM AUTHOR]

Writer identification has steadily progressed in recent decades owing to its widespread application. Scenarios with extensive handwriting data such as page-level or sentence-level have achieved satisfactory accuracy; however, word-level offline writer identification is still challenging owing to the difficulty of learning good feature representations with scant handwriting data. This paper proposes a new Residual Swin Transformer Classifier (RSTC), which comprehensively aggregates local and global handwriting styles and yields robust feature representations with single-word images. The local information is modeled by the Transformer Block through interacting strokes and global information is featurized by holistic encoding using the Identity Branch and Global Block. Moreover, the pre-training technique is exploited to transfer reusable knowledge learned from a task similar to writer identification, strengthening RSTC’s representation of handwriting features. The proposed method is tested on the IAM and CVL benchmark datasets and achieves state-of-the-art performance, which demonstrates the superior modeling capability of RSTC for word-level writer identification.

Writer identification has become a hot research topic in the fields of pattern recognition, forensic document analysis, the criminal justice system, etc. The goal of this research is to propose an efficient approach for writer identification based on online handwritten Kanji characters. We collected 47,520 samples from 33 people who wrote 72 online handwritten-based Kanji characters 20 times. We extracted features from the handwriting data and proposed a support vector machine (SVM)-based classifier for writer identification. We also conducted experiments to see how the accuracy changes with feature selection and parameter tuning. Both text-dependent and text-independent writer identification were studied in this work. In the case of text-dependent writer identification, we obtained the accuracy of each Kanji character separately. We then studied the text-independent case by considering some of the top discriminative characters from the text-dependent case. Finally, another text-dependent experiment was performed by taking two, three, and four Kanji characters instead of using only one character. The experimental results illustrated that SVM provided the highest identification accuracy of 99.0% for the text-independent case and 99.6% for text-dependent writer identification. We hope that this study will be helpful for writer identification using online handwritten Kanji characters. [ABSTRACT FROM AUTHOR]

Offline handwriting has become a very important part in the field of behavioral biometrics. In this paper, we proposed an approach to identify writers from their handwritten documents. The foremost contribution of this study is to suggest the use of co-occurrence features to improve the performance of writer identification. A contour texture-based feature is extracted from preprocessed regions of interest (components or sub-images) and their exterior contours are based primarily on Modified Local Binary Pattern (MLBP) and Ink-trace Width and Shape Letters (IWSL) measurements. Considering the contour as a texture, and using these textural descriptors, the joint probability distributions of MLBP and IWSL on different pixels are calculated in order to determine the similarities between different images of handwriting. Identification is carried out using the nearest neighbor rule and Chi-square distance. The proposed system has been evaluated on eight well-known handwriting databases (Arabic IFN/ENIT and KHATT, English IAM and CVL, Dutch Firemaker, Portuguese BFL, Chinese CERUG-CN, and English/Greek ICDAR2013). Experimental results show that the recommended scheme achieves the highest performance on KHATT, CVL, Firemaker, BFL, CERUG-CN, and ICDAR2013 databases, and that it demonstrates competitive performance on IFN/ENIT and IAM databases as compared to those reported by the state-of-the-art identification systems. [ABSTRACT FROM AUTHOR]

Determining the writer or transcriber of historical Arabic manuscripts has always been a major challenge for researchers in the field of humanities. With the development of advanced techniques in pattern recognition and machine learning, these technologies have been applied to automate the extraction of paleographical features in order to solve this issue. This paper presents a baseline system for writer identification, tested on a Historical Arabic dataset of 11610 single and double folio images. These texts were extracted from a unique collection of 567 Historical Arabic Manuscripts available at the Balamand Digital Humanities Center. A survey has been conducted on the available Arabic datasets and previously proposed techniques and algorithms. The Balamand dataset presents an important challenge due to the geo-historical identity of manuscripts and their physical conditions. An advanced Deep Learning system was developed and tested on three different Latin and Arabic datasets: ICDAR19, ICFHR20 and KHATT, before testing it on the Balamand dataset. The system was compared with many other systems and it has yielded a state-of-the-art performance on the new challenging images with 95.2% mean Average Precision (mAP) and 98.1% accuracy. [ABSTRACT FROM AUTHOR]