A Sequential Handwriting Recognition Model Based on a Dynamically Configurable CRNN

Al-Saffar, Ahmed; Awang, Suryanti; Al-Saiagh, Wafaa; Al-Khaleefa, Ahmed Salih; Abed, Saad Adnan

doi:10.3390/s21217306

Cited by 12 publications

(4 citation statements)

References 67 publications

(85 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Future integration of varied analysis methods for the handwriting data taken from handwriting recognition methods [45], machine learning algorithms [46], and combing intelligent pen devices [47] may support the PD-MCI early detection process.…”

Section: Discussionmentioning

confidence: 99%

Patients’ Self-Report and Handwriting Performance Features as Indicators for Suspected Mild Cognitive Impairment in Parkinson’s Disease

Rosenblum

Meyer

Richardson

et al. 2022

Sensors

View full text Add to dashboard Cite

Early identification of mild cognitive impairment (MCI) in Parkinson’s disease (PD) patients can lessen emotional and physical complications. In this study, a cognitive functional (CF) feature using cognitive and daily living items of the Unified Parkinson’s Disease Rating Scale served to define PD patients as suspected or not for MCI. The study aimed to compare objective handwriting performance measures with the perceived general functional abilities (PGF) of both groups, analyze correlations between handwriting performance measures and PGF for each group, and find out whether participants’ general functional abilities, depression levels, and digitized handwriting measures predicted this CF feature. Seventy-eight participants diagnosed with PD by a neurologist (25 suspected for MCI based on the CF feature) completed the PGF as part of the Daily Living Questionnaire and wrote on a digitizer-affixed paper in the Computerized Penmanship Handwriting Evaluation Test. Results indicated significant group differences in PGF scores and handwriting stroke width, and significant medium correlations between PGF score, pen-stroke width, and the CF feature. Regression analyses indicated that PGF scores and mean stroke width accounted for 28% of the CF feature variance above age. Nuances of perceived daily functional abilities validated by objective measures may contribute to the early identification of suspected PD-MCI.

show abstract

Section: Discussionmentioning

confidence: 99%

Patients’ Self-Report and Handwriting Performance Features as Indicators for Suspected Mild Cognitive Impairment in Parkinson’s Disease

Rosenblum

Meyer

Richardson

et al. 2022

Sensors

View full text Add to dashboard Cite

show abstract

“…The CRNN is composed of three components: a convolutional layer, a recurrent layer, and a CTC layer. Saffar et al proposed to use the salp swarm optimization algorithm to optimize the parameters of convolutional neural network in DC-CRNN [ 48 ] to further improve the recognition accuracy of CRNN. The multi-modal text recognition network (MATRN) [ 49 ] proposed by Na et al can better improve the accuracy of text recognition by fusing visual and semantic information.…”

Section: Related Workmentioning

confidence: 99%

Text Recognition Model Based on Multi-Scale Fusion CRNN

Zou

Wang

et al. 2023

Sensors

View full text Add to dashboard Cite

Scene text recognition is a crucial area of research in computer vision. However, current mainstream scene text recognition models suffer from incomplete feature extraction due to the small downsampling scale used to extract features and obtain more features. This limitation hampers their ability to extract complete features of each character in the image, resulting in lower accuracy in the text recognition process. To address this issue, a novel text recognition model based on multi-scale fusion and the convolutional recurrent neural network (CRNN) has been proposed in this paper. The proposed model has a convolutional layer, a feature fusion layer, a recurrent layer, and a transcription layer. The convolutional layer uses two scales of feature extraction, which enables it to derive two distinct outputs for the input text image. The feature fusion layer fuses the different scales of features and forms a new feature. The recurrent layer learns contextual features from the input sequence of features. The transcription layer outputs the final result. The proposed model not only expands the recognition field but also learns more image features at different scales; thus, it extracts a more complete set of features and achieving better recognition of text. The results of experiments are then presented to demonstrate that the proposed model outperforms the CRNN model on text datasets, such as Street View Text, IIIT-5K, ICDAR2003, and ICDAR2013 scenes, in terms of text recognition accuracy.

show abstract

“…The subsequent step is creating a recurrent network, which is responsible for making frame-to-frame predictions. It is important for this process to be complete because it is the source of the model's accuracy (Al-Saffar et al, 2021). Although a CRNN consists of two separate network topologies (a DCNN and an RNN), it is feasible to train it concurrently using a single loss function.…”

Section: Overall Approachmentioning

confidence: 99%

An empirical study of extracting embedded text from digital images

Shafie

2023

Int. j. adv. appl. sci.

View full text Add to dashboard Cite

The utilization of images as a means of transferring information is a widespread technique employed to circumvent simple detection functions that primarily focus on analyzing textual content rather than conducting thorough file examinations. This study investigates the efficacy of deep learning models in detecting embedded information within digital images. The data used for analysis was acquired from a secondary source and underwent comprehensive preprocessing. Feature extraction, sequence labeling, and predictive model training were performed using CRNN, CNN, and RNN models. Two specific models were trained and tested in this research: 1) CNN, RNN-LSTM with the Adam optimizer, and 2) CNN, RNN-GRU with the RAdam optimizer for text detection. The findings reveal that Model #1 achieved the highest F1-score during testing, with a score of 98.37% for text detection and 96.73% for word detection. The second model obtained an F1-score of 94.84% and 93.05% for text and word detection, respectively. Model #1 exhibited a word detection accuracy of 98.38% and a text detection accuracy of 96.47%. These findings indicate that the first model outperformed the second model, suggesting that employing RNN-LSTM and the Adam optimizer made a positive impact. Therefore, utilizing deep learning tools and emerging technologies is crucial for extracting textual information and analyzing visual data. In summary, this study concludes that deep learning models can be relied upon to effectively detect textual information embedded within digital images.

show abstract

A Sequential Handwriting Recognition Model Based on a Dynamically Configurable CRNN

Cited by 12 publications

References 67 publications

Patients’ Self-Report and Handwriting Performance Features as Indicators for Suspected Mild Cognitive Impairment in Parkinson’s Disease

Patients’ Self-Report and Handwriting Performance Features as Indicators for Suspected Mild Cognitive Impairment in Parkinson’s Disease

Text Recognition Model Based on Multi-Scale Fusion CRNN

An empirical study of extracting embedded text from digital images

Contact Info

Product

Resources

About