Online Handwritten Research Articles

Explainability of CNN-based Alzheimer's disease detection from online handwriting.

With over 55 million people globally affected by dementia and nearly 10 million new cases reported annually, Alzheimer's disease is a prevalent and challenging neurodegenerative disorder. Despite significant advancements in machine learning techniques for Alzheimer's disease detection, the widespread adoption of deep learning models raises concerns about their explainability. The lack of explainability in deep learning models for online handwriting analysis is a critical gap in the literature in the context of Alzheimer's disease detection. This paper addresses this challenge by interpreting predictions from a Convolutional Neural Network applied to multivariate time series data, generated by online handwriting data associated with continuous loop series handwritten on a graphical tablet. Our explainability methods reveal distinct motor behavior characteristics for healthy individuals and those diagnosed with Alzheimer's. Healthy subjects exhibited consistent, smooth movements, while Alzheimer's patients demonstrated erratic patterns marked by abrupt stops and direction changes. This emphasizes the critical role of explainability in translating complex models into clinically relevant insights. Our research contributes to the enhancement of early diagnosis, providing significant and reliable insights to stakeholders involved in patient care and intervention strategies. Our work bridges the gap between machine learning predictions and clinical insights, fostering a more effective and understandable application of advanced models for Alzheimer's disease assessment.

A multi-modal Parkinson’s disease diagnosis system from EEG signals and online handwritten tasks using grey wolf optimization based deep learning model

Parkinson’s disease (PD) is a neurodegenerative disorder characterized by the gradual deterioration of motor function, affecting speech, writing, muscle control, and mobility. The existing studies have not utilized both the electroencephalography (EEG) signals and online handwritten tasks together to diagnose PD. The studies have also not explored the EEG signals collected from specific brain regions like the substansia niagra (SN) and ventral tegmental area (VTA), crucial for dopamine production linked to PD. This article proposes a multi-modal PD diagnosis system from EEG signals (collected from SN and VTA regions of the brain), collected during performing online handwritten tasks, using grey wolf optimization (GWO) algorithm. Mel-frequency cepstral coefficients (MFCC) features have been generated from the EEG signals and optimized by the GWO algorithm. The classification (diagnosis) experiments on the optimal number of feature values, obtained from GWO algorithm, have been carried out using bidirectional long short-term memory (BLSTM) variant of recurrent neural network (RNN). The classification experiments have also been conducted using support vector machine (SVM), bagged random forest (BRF), and long short-term memory (LSTM) variant of RNN classifier to have a performance comparison with the proposed method. The effectiveness of the introduced PD diagnosis system has been analyzed on a self-generated dataset named EEG signal based on online handwriting (ESOH). A maximum classification accuracy of 99.30% has been achieved from the proposed PD diagnosis system. The experimental outcomes illustrate that the introduced PD diagnosis system outperforms the state-of-the-art PD diagnosis systems relying on the EEG signals for diagnosing the PD.

Increasing Offline Handwritten Chinese Character Recognition Using Separated Pre-Training Models: A Computer Vision Approach

Offline handwritten Chinese character recognition involves the application of computer vision techniques to recognize individual handwritten Chinese characters. This technology has significantly advanced the research in online handwriting recognition. Despite its widespread application across various fields, offline recognition faces numerous challenges. These challenges include the diversity of glyphs resulting from different writers’ styles and habits, the vast number of Chinese character labels, and the presence of morphological similarities among characters. To address these challenges, an optimization method based on a separated pre-training model was proposed. The method aims to enhance the accuracy and robustness of recognizing similar character images by exploring potential correlations among them. In experiments, the HWDB and Chinese Calligraphy Styles by Calligraphers datasets were employed, utilizing precision, recall, and the Macro-F1 value as evaluation metrics. We employ a convolutional self-encoder model characterized by high recognition accuracy and robust performance. The experimental results demonstrated that the separated pre-training models improved the performance of the convolutional auto-encoder model, particularly in handling error-prone characters, resulting in an approximate 6% increase in precision.

A CNN-KNN Based Recognition of Online Handwritten Symbols within Physics Expressions Using Contour-Based Bounding Box (CBBS) Segmentation Technique

A CNN-KNN Based Recognition of Online Handwritten Symbols within Physics Expressions Using Contour-Based Bounding Box (CBBS) Segmentation Technique

Classifying of Isolated Handwritten Arabic Characters (on-line) Using Machine Learning Methods

The prediction of Arabic handwritten characters is one of the fascinating matters in the area of artificial intelligence (AI) and machine learning, especially in the case of on-line handwriting. In the handwriting recognition system, characters are classified according to their particular categories by utilizing -as input- the values of the features extracted from them. Our major objective in this paper is to classify the isolated characters for online Arabic handwriting (? to ?) using machine learning techniques with the features from EDMs (Edge Direction Matrixes) as opposed to EEDMs (Extended Edge Direction Matrixes). With respect to all classification methods used in this study, EEDMs with machine learning techniques had better performance than EDMs in recognizing isolated Arabic online characters.

A self-attention-based deep architecture for online handwriting recognition

A self-attention-based deep architecture for online handwriting recognition

Advanced Recognizing and Categorizing System for Alzheimer's and Parkinson's Syndromes using CNN Algorithm and Clinical Data

Parkinson’s Disease (PD) and Alzheimer’s Disease (AD) are advancing neurodegenerative diseases that impact a large number of individuals globally. This study provides a novel method to identify and categorize individuals in the early stages of PD and AD by analyzing online handwriting structures, leveraging the current progress in Machine Learning (ML) in the medical domain. The research explores multiple data augmentation strategies to address the limited training information available for classifying neuro-degenerative disorders based on behavioral information. The study used a Convolutional Neural Network (CNN) classification as a foundation model by identifying essential radiomics characteristics from the T1-weighted Magnetic Resonance Imaging (MRI) data. The research explores data augmentation strategies designed explicitly for time-series information, such as online handwriting patterns, instead of standard methods used for hardware-based detection. This approach, with more validation, might assist in distinguishing between difficult instances of AD and PD that have a comparable mild motor and non-motor symptoms.

Handwritten online signature verification and Forgery Detection for Online Handwritten Signature using Hybrid Wavelet Transform-1 with HMM classifier

Online signature verification employs a distinctive biometric trait by utilizing both static and dynamic features extracted from 2D signature images. A hybrid wavelet transform, denoted as HWT-1 with a size of 256, is formed through the Kronecker product of two orthogonal transforms, such as DCT, DHT, Haar, Hadamard, and Kekre, each with sizes 4 and 64. This HWT facilitates signal analysis at both global and local levels, akin to traditional wavelet transforms. Specifically, HWT-1 processes the 256 samples of online handwritten signatures, yielding 128 samples that constitute the feature vectors for signature verification and forgery detection. These feature vectors are then inputted into the Left-Right and Ergodic Hidden Markov Model (HMM) classifiers for further analysis. The HMMs are trained using 10 randomly selected genuine signature samples and subsequently tested with the remaining 10 genuine signatures and 20 forged signatures from the SVC 2004 signature database, repeating this process 20 times to compute average values. Among all possible combinations of HWT-1 utilizing DCT, DHT, Haar, Hadamard, and Kekre transforms for the Left-Right HMM model, the combination of DCT 4 and Haar 64 demonstrates the best performance with a False Rejection Rate (FRR) and False Acceptance Rate (FAR) of 1.05% and 0.99%, respectively, at state 3. Similarly, considering all feasible combinations of HWT-1 for the Ergodic HMM model, the combination of DCT 4 and Kekre 64 yields the optimal performance with an FRR and FAR of 1.24% and 1.33%, respectively, at state 3.

Handwritten Online Signature Verification and Forgery Detection Using Hybrid Wavelet Transform-2 with HMM Classifier

Online signature verification stands out as a distinctive biometric feature, offering both static and dynamic attributes within 2D signature images. A Hybrid Wavelet Transform-2 (HWT-2) with a size of 256 is constructed by employing the Kronecker product of two orthogonal transforms: DCT, DHT, Haar, Hadamard, and Kekre, each with sizes of 4 and 64. The HWT enables the analysis of signals at both global and local levels, akin to wavelet transforms. HWT-2 is applied to 256 samples of online handwritten signatures, and the first 128 samples of the output are utilized as feature vectors for the verification and forgery detection of online handwritten signatures. These feature vectors are inputted into Left-Right and Ergodic Hidden Markov Model (HMM) classifiers for analysis. The HMMs are trained using 10 randomly selected genuine signature samples, and subsequently tested on the remaining 10 genuine signatures and 20 forged signatures from 40 users of the SVC 2004 signature database. This process is repeated 20 times, and the average values are computed. Among all possible combinations of HWT-2 using DCT, DHT, Haar, Hadamard, and Kekre transforms for the Left-Right HMM model, the combination of DCT 4 and DHT 64 demonstrates the best performance, with False Rejection Rate (FRR) and False Acceptance Rate (FAR) values of 3.96% and 1.48%, respectively, for state 5. Similarly, for the Ergodic HMM model, the combination of DCT 4 and DHT 64 exhibits the best performance, with FRR and FAR values of 1.10% and 2.88%, respectively, for state 5. These results indicate that combinations of HWT-2 outperform individual orthogonal transforms, and further, that HWT-2 combinations within the Ergodic HMM model offer superior performance compared to the Left-Right HMM model.

SpaceGTN: A Time-Agnostic Graph Transformer Network for Handwritten Diagram Recognition and Segmentation

Online handwriting recognition is pivotal in domains like note-taking, education, healthcare, and office tasks. Existing diagram recognition algorithms mainly rely on the temporal information of strokes, resulting in a decline in recognition performance when dealing with notes that have been modified or have no temporal information. The current datasets are drawn based on templates and cannot reflect the real free-drawing situation. To address these challenges, we present SpaceGTN, a time-agnostic Graph Transformer Network, leveraging spatial integration and removing the need for temporal data. Extensive experiments on multiple datasets have demonstrated that our method consistently outperforms existing methods and achieves state-of-the-art performance. We also propose a pipeline that seamlessly connects offline and online handwritten diagrams. By integrating a stroke restoration technique with SpaceGTN, it enables intelligent editing of previously uneditable offline diagrams at the stroke level. In addition, we have also launched the first online handwritten diagram dataset, OHSD, which is collected using a free-drawing method and comes with modification annotations.

Parkinson’s Disease Detection From Online Handwriting Based on Beta-Elliptical Approach and Fuzzy Perceptual Detector

Parkinson’s Disease Detection From Online Handwriting Based on Beta-Elliptical Approach and Fuzzy Perceptual Detector

Online Handwritten Chinese Character Recognition Based on 1-D Convolution and Two-Streams Transformers

Online Handwritten Chinese Character Recognition Based on 1-D Convolution and Two-Streams Transformers

Handwriting Enhancement: Recognition-Based and Recognition-Independent Approaches for On-Device Online Handwritten Text Alignment

Handwriting Enhancement: Recognition-Based and Recognition-Independent Approaches for On-Device Online Handwritten Text Alignment

Transformer-based stroke relation encoding for online handwriting and sketches

Stroke classification for online handwriting and sketches, aimed at grouping strokes into different semantic categories, has drawn considerable attention due to its wide applications. This task is challenging since individual strokes look similar and are easily confused with each other. The key is to consider both the individual strokes and the contextual information jointly for making prediction. Previous methods are insufficient in modeling and exploiting complex contextual information of strokes. To overcome this limitation, we propose a Transformer-based model for Online Handwriting and Sketches (T-OHS), with novel relation encoding schemes to take advantage of temporal and spatial information in stroke sequence. Particularly, we introduce a coarse-to-fine hierarchical encoding approach based on the polar coordinate system for precisely modeling spatial relations between strokes. Experiments on three types of handwriting data, including online handwritten documents, diagrams, and sketches, demonstrate that our method is universal and provides state-of-the-art performance.

Online Handwritten Signature Verification Method Based on Uni-Feature Correlation Coefficient between Signatures

Online handwritten signature verification is a crucial direction of research in the field of biometric recognition. Recently, many studies concerning online signature verification have attempted to improve performance using multi-feature fusion. However, few studies have provided the rationale for selecting a certain uni-feature to be fused, and few studies have investigated the contributions of a certain uni-feature in the multi-feature fusion process. This lack of research makes it challenging for future researchers in related fields to gain inspiration. Therefore, we use the uni-feature as the research object. In this paper, the uni-feature is one of the X and Y coordinates of the signature trajectory point, pen pressure, pen tilt, and pen azimuth feature. Aiming to solve the unequal length of feature vectors and the low accuracy of signature verification when using uni-features, we innovatively introduced the idea of correlation analysis and proposed a dynamic signature verification method based on the correlation coefficient of uni-features. Firstly, an alignment method of two feature vector lengths was proposed. Secondly, the correlation coefficient calculation formula was determined by analyzing the distribution type of the feature data, and then the correlation coefficient of the same uni-feature between the genuine signatures or between the genuine and forged signatures was calculated. Finally, the signature was verified by introducing a Gaussian density function model and combining it with the signature verification discrimination threshold. Experimental results showed that the proposed method could improve the performance of dynamic signature verification based on uni-features. In addition, the pen pressure feature had the best signature verification performance, with the highest signature verification accuracy of 93.46% on the SVC 2004 dataset.

BWordDeepNet: a novel deep learning architecture for the recognition of online handwritten Bangla words

BWordDeepNet: a novel deep learning architecture for the recognition of online handwritten Bangla words

A Study of Techniques and Challenges in Text Recognition Systems

The core system for Natural Language Processing (NLP) and digitalization is Text Recognition. These systems are critical in bridging the gaps in digitization produced by non-editable documents, as well as contributing to finance, health care, machine translation, digital libraries, and a variety of other fields. In addition, as a result of the pandemic, the amount of digital information in the education sector has increased, necessitating the deployment of text recognition systems to deal with it. Text Recognition systems worked on three different categories of text: (a) Machine Printed, (b) Offline Handwritten, and (c) Online Handwritten Texts. The major goal of this research is to examine the process of typewritten text recognition systems. The availability of historical documents and other traditional materials in many types of texts is another major challenge for convergence. Despite the fact that this research examines a variety of languages, the Gurmukhi language receives the most focus. This paper shows an analysis of all prior text recognition algorithms for the Gurmukhi language. In addition, work on degraded texts in various languages is evaluated based on accuracy and F-measure.

Parkinson’s disease diagnosis using recurrent neural network based deep learning model by analyzing online handwriting

Parkinson’s disease diagnosis using recurrent neural network based deep learning model by analyzing online handwriting

Multimodal Biometric System Based on the Fusion in Score of Fingerprint and Online Handwritten Signature

Abstract Multimodal biometrics is the technique of using multiple modalities on a single system. This allows us to overcome the limitations of unimodal systems, such as the inability to acquire data from certain individuals or intentional fraud, while improving recognition performance. In this paper, a study of score normalization and its impact on the performance of the system is performed. The fusion of scores requires prior normalisation before applying a weighted sum fusion that separates impostor and genuine scores into a common interval with close ranges. The experiments were carried out on three biometric databases. The results show that the proposed strategy performs very encouragingly, especially in combination with Empirical Modal Decomposition (EMD). The proposed fusion system shows good performance. The best result is obtained by merging the globality online signature and fingerprint where an EER of 1.69 % is obtained by normalizing the scores according to the Min-Max method.

Off-Line and Online Handwritten Character Recognition Using RNN-GRU Algorithm

Abstract: Recognizing handwritten characters is an extremely difficult task in the domains of pattern recognition and computer vision. It involves the use of a process that enables computers to identify and convert handwritten or printed characters, such as letters and numbers, into a digital format that is usable by the computer. Currently, the RNN-CNN hybrid algorithm is employed to predict handwritten text in images with an accuracy rate of 91.5%. However, the existing system can only recognize characters and words character-by-character and word-by-word. The proposed system aims to address this limitation by enabling line-byline recognition and the conversion of handwritten text to OCR. To achieve this, the system utilizes the GRU algorithm to predict the next letter in incomplete words. Furthermore, the IAM dataset, consisting of 135,000 annotated sentences, is utilized to detect and rectify spelling errors in texts.