Character Recognition Research Articles

Integrating YOLOv8 and CSPBottleneck based CNN for enhanced license plate character recognition

Integrating YOLOv8 and CSPBottleneck based CNN for enhanced license plate character recognition

Intramolecular hydrogen bonding steered optical recognition of fluoride ion and consequent opto-chemical logic responses: Spectroscopic and computational studies

This work demonstrates the synthesis and characterization of a thiosemicarbazide core containing Schiff base molecule (MNHC) and its use as a chemosensor for the selective, ratiometric, and colorimetric detection of fluoride ions. The study reveals that in the presence of F− ions, MNHC displays a ratiometric change of absorption maxima at two different wavelengths (432 and 330 nm). MNHC involves an intramolecular hydrogen bonding initiating the possibility of keto-enol tautomerism which is thoroughly investigated by DFT study. Interestingly, the intramolecular hydrogen bonding enhances the acidity of the –NH proton and consequently the –NH center becomes more labile for F− ions binding over the other –NH proton. This fact is confirmed by 1H NMR study. Remarkably, F− ion forms intermolecular hydrogen bonding with the –NH moiety disrupting the intramolecular hydrogen bonding. The changes in absorption spectra in the presence of F− ions occur due to the intermolecular hydrogen bonding-assisted abstraction of the –NH proton. On the other hand, when H2O is added to the MNHC-F− composite, the spectroscopic signature of free MNHC is completely regenerated due to the transfer of proton from H2O to the nitrogen center of MNHC. In addition, 1:1 binding stoichiometry and 53.7 μM detection limit for fluoride ion are also confirmed. Moreover, utilizing these spectroscopic responses, different kinds of important opto-chemical logic gates like YES, NOT, PASS 0, PASS 1, INHIBIT, TRANSFER, and IMPLICATION are developed. Moreover, the F− ion-assisted competitive nature between the intramolecular vs intermolecular hydrogen bonding is thoroughly explored by substantial DFT study.

An End-to-End Scene Text Recognition for Bilingual Text

Text localization and recognition from natural scene images has gained a lot of attention recently due to its crucial role in various applications, such as autonomous driving and intelligent navigation. However, two significant gaps exist in this area: (1) prior research has primarily focused on recognizing English text, whereas Arabic text has been underrepresented, and (2) most prior research has adopted separate approaches for scene text localization and recognition, as opposed to one integrated framework. To address these gaps, we propose a novel bilingual end-to-end approach that localizes and recognizes both Arabic and English text within a single natural scene image. Specifically, our approach utilizes pre-trained CNN models (ResNet and EfficientNetV2) with kernel representation for localization text and RNN models (LSTM and BiLSTM) with an attention mechanism for text recognition. In addition, the AraElectra Arabic language model was incorporated to enhance Arabic text recognition. Experimental results on the EvArest, ICDAR2017, and ICDAR2019 datasets demonstrated that our model not only achieves superior performance in recognizing horizontally oriented text but also in recognizing multi-oriented and curved Arabic and English text in natural scene images.

Simple techniques to bypass GenAI text detectors: implications for inclusive education

This study investigates the efficacy of six major Generative AI (GenAI) text detectors when confronted with machine-generated content modified to evade detection (n = 805). We compare these detectors to assess their reliability in identifying AI-generated text in educational settings, where they are increasingly used to address academic integrity concerns. Results show significant reductions in detector accuracy (17.4%) when faced with simple techniques to manipulate the AI generated content. The varying performances of GenAI tools and detectors indicate they cannot currently be recommended for determining academic integrity violations due to accuracy limitations and the potential for false accusation which undermines inclusive and fair assessment practices. However, these tools may support learning and academic integrity when used non-punitively. This study aims to guide educators and institutions in the critical implementation of AI text detectors in higher education, highlighting the importance of exploring alternatives to maintain inclusivity in the face of emerging technologies.

Standards-based character recognition test

Abstract The present study developed and implemented a character recognition test based on the Chinese Proficiency Grading Standards for International Chinese Language Education. Using a stratified systematic sampling approach, we selected 15 characters from each level, ranging from levels 1 to 6, and 45 characters from the advanced level. This test was administered to 32 L2 Chinese learners from two universities. The results, on one hand, present an estimation of learners’ character size and revealed variations at multiple levels, including school, course, and individual levels. On the other hand, they serve as a valuable link between the Standards and teaching/learning practice, enabling a necessary reflection on the efficacy of the Standards, particularly in terms of the character component.

ESRNet: an exploring sample relationships network for arbitrary-shaped scene text detection

ESRNet: an exploring sample relationships network for arbitrary-shaped scene text detection

Bangla Optical Character Recognition for Mobile Platforms: A Comprehensive Cross-Platform Approach

The development of Optical Character Recognition (OCR) systems for Bangla script has been an area of active research since the 1980s. This study presents a comprehensive analysis and development of a cross-platform mobile application for Bangla OCR, leveraging the Tesseract OCR engine. The primary objective is to enhance the recognition accuracy of Bangla characters, achieving rates between 90% and 99%. The application is designed to facilitate the automatic extraction of text from images selected from the device's photo library, promoting the preservation and accessibility of Bangla language materials. This paper discusses the methodology, including the preparation of training datasets, preprocessing steps, and the integration of the Tesseract OCR engine within a Dart programming environment for cross-platform functionality. This integration provides that the application could be introduced on mobile platforms without substantial alterations. The results demonstrate significant improvements in recognition accuracy, making this application a valuable tool for various practical applications such as data entry for printed Bengali documents, automatic recognition of Bangla number plates, and the digital archiving of vintage Bangla books. These improvements are crucial to further enhance the usability and reliability of Bangla OCR on mobile devices. Our cross-platform method for Bangla OCR on mobile devices provides a strong solution with exceptional identification accuracy, which helps in preserving and making Bangla language information accessible in digital format. This study has significant implications for future research and advancement in the field of optical character recognition (OCR) for intricate writing systems, especially in mobile settings.

Rethinking Multilingual Scene Text Spotting: A Novel Benchmark and a Character-Level Feature Based Approach

End-to-end multilingual scene text spotting aims to integrate scene text detection and recognition into a unified framework. Actually, the accuracy of text recognition largely depends on the accuracy of text detection. Due to the lackage of benchmarks with adequate and high-quality character-level annotations for multilingual scene text spotting, most of the existing methods train on the benchmarks only with word-level annotations. However, the performance of multilingual scene text spotting are not that satisfied training on the existing benchmarks, especially for those images with special layout or words out of vocabulary. In this paper, we proposed a simple YOLO-like baseline named CMSTR for character-level multilingual scene text spotting simultaneously and efficiently. Technically, for each text instance, we represent the character sequence as ordered points and model them with learnable explicit point queries. After passing a single decoder, the point queries have encoded requisite text semantics and locations, thus can be further decoded to the center line, boundary, script, and confidence of text via very simple prediction heads in parallel. Furthermore, we show the surprisingly good extensibility of our method, in terms of character class, language type, and task. On the one hand, DeepSolo not only performs well in English scenes but also masters the Chinese transcription with complex font structure and a thousand-level character classes. On the other hand, based on the extensibility of DeepSolo, we launch DeepSolo++ for multilingual text spotting, making a further step to let Transformer decoder with explicit points solo for multilingual text detection, recognition, and script identification all at once.

An open dataset for oracle bone character recognition and decipherment

Oracle bone script, one of the earliest known forms of ancient Chinese writing, presents invaluable research materials for scholars studying the humanities and geography of the Shang Dynasty, dating back 3,000 years. The immense historical and cultural significance of these writings cannot be overstated. However, the passage of time has obscured much of their meaning, presenting a significant challenge in deciphering these ancient texts. With the advent of Artificial Intelligence (AI), employing AI to assist in deciphering Oracle Bone Characters (OBCs) has become a feasible option. Yet, progress in this area has been hindered by a lack of high-quality datasets. To address this issue, this paper details the creation of the HUST-OBC dataset. This dataset encompasses 77,064 images of 1,588 individual deciphered characters and 62,989 images of 9,411 undeciphered characters, with a total of 140,053 images, compiled from diverse sources. The hope is that this dataset could inspire and assist future research in deciphering those unknown OBCs. All the codes and datasets are available at https://github.com/Pengjie-W/HUST-OBC.

BEYOND THE ROLLCALL: SMART STRATEGY FOR ATTENDANCE

Attendance tracking is a crucial aspect of educational institutions, yet the traditional manual methods are time-consuming and error-prone. This project presents an innovative solution to this problem through the development of an automated attendance system. Utilizing advanced technologies such as face detection and optical character recognition (OCR), this system aims to streamline the attendance tracking process in schools and colleges. By automatically detecting student faces and extracting roll numbers from stickers affixed to the left side of their faces, the system eliminates the need for manual intervention, thereby improving efficiency and accuracy. Additionally, a user-friendly GUI created using customtkinter in Python enhances the system's usability and accessibility. Attendance records are automatically stored in an Excel file, allowing for easy tracking and analysis by educators.

HELMET DETECTION AND LICENSE PLATE RECOGNITION USING CNN

Right now, we're facing a lot of issues with traffic rules in India that might use some fresh perspectives to fix. There has been a rise in the number of accidents and fatalities in India caused by the traffic offense of riding motorcycles or mopeds without a helmet. The current system mostly relies on CCTV records to track traffic offenses. In such cases, the traffic police have to zoom in on the license plate to identify the offending rider if they aren't wearing a helmet. The traffic offenses are common, and the number of persons riding motorbikes is growing daily, so this takes a lot of time and effort. Imagine a system that could detect whether a motorcyclist or moped rider isn't wearing a helmet and, if found, immediately get the license plate number. Recent studies have effectively accomplished this task using various characteristics such as CNN, R-CNN, LBP, HoG, HaaR, etc. In terms of speed, accuracy, and efficiency, however, these works have their limitations when it comes to object recognition and categorization. To try to automate the process of finding drivers who don't wear helmets and getting their license plate numbers, this study developed a system called Non-Helmet Rider detection. The core idea is based on three-level deep learning for object detection. Using YOLOv2, the first level detects a person and a motorbike or moped; the second level uses YOLOv3, and the third level uses YOLOv2. The items recognized are a helmet and a license plate. After then, Optical Character Recognition is used to obtain the license plate registration number. There are some limitations and requirements placed on all of these methods, particularly the one that extracts license plates. The efficiency of the process is crucial since video is used as an input in this task. We have developed a comprehensive system that can recognize helmets and retrieve license plate numbers using the aforementioned approaches.

A revisit to interactions of stimulus quality and semantic context on N400 in visual word recognition.

Much behavioral research has revealed interactive effects between stimulus quality and semantic priming in visual word recognition, practically in favor of the interactive activation model. However, the limited number of event-related brain potential (ERP) studies have yielded inconsistent results considering this interaction's impact on N400 amplitude. The current ERP study aimed to examine whether the joint effects of stimulus quality and semantic priming were specific to the lexical decision task. We used both behavioral measures and ERP recordings to evaluate the joint effects of stimulus degradation (i.e. highly vs. slightly degraded) and semantic priming (i.e. semantically related vs. unrelated) in a lexical decision task involving visual recognition of Chinese characters. The results showed significant degradation-by-priming interactions on response times and N400 amplitude ( P < 0.05), with larger semantic priming effects on slightly degraded targets. These converging behavioral and electrophysiological findings provide evidence in accordance with the interactive activation models of visual word recognition, in which the early-stage visual processing (i.e. degradation) cascades into the later-stage semantic processing (i.e. priming), thus yielding interactions observed in N400 amplitude.

MODEL IMPLEMENTATION AND COMPREHENSIVE STUDY ON VISUAL AIR TYPING SYSTEMS

This paper explores the implementation and applications of visual air typing systems, focusing on fingertip detection and text recognition to provide an alternative to the traditional input methods. Although models with similar functionality exist, our study emphasizes how such systems can offer practical solutions for users with visual impairments and enhance user experience in virtual reality (VR) environments. By detecting fingertip movements and translating them into written words, such models present a seamless and intuitive user experience. For individuals with visual impairments, traditional keyboards can be challenging, and sign language lacks uniformity across different regions. An air typing system addresses these issues by offering a consistent text input method. Additionally, in VR applications where traditional keyboards are impractical, it enables more natural and efficient text entry. This paper provides a comprehensive review of existing air typing technologies, detailing their strengths and limitations. We discuss the potential future directions for air typing systems, highlighting the need for more inclusive and versatile input methods. Our findings indicate that air typing can significantly enhance user interaction in diverse environments, making it a valuable alternative to traditional keyboards. Overall, this study underscores the practical applications and benefits of air typing systems, particularly for visually impaired users and VR contexts, contributing to advancing more inclusive technological solutions.

Crowdsourcing the digital parliament

ABSTRACT Parliaments will eventually not evade the digital evolution of every institution to become data-driven organisations. This study examines the Hellenic Optical Character Recognition (OCR) Team, a pioneering crowdsourcing initiative aimed at processing and analysing parliamentary data. Under certain conditions, crowdsourcing, in other words, the power of the people, can be appropriately channelled and exploited to support representative institutions and their societal stakeholders in managing their change processes. Based on survey findings, this research highlights the profiles and motivations of participants, identifies best practices for crowdsourcing in parliamentary contexts, and demonstrates the initiative's impact on improving data accessibility and transparency. The results suggest that the Hellenic OCR Team significantly contributes to the digital evolution of parliamentary functions and promotes greater public involvement in the legislative process.

Esmeraldino Salles (1916 - 1979) documentado pela imprensa: uma incursão no Acervo da Hemeroteca Digital da Biblioteca Nacional

This article describes the process, methodology and results of a search in the National Library’s Digital Library for the São Paulo musician Esmeraldino Salles, who lived between 1916 and 1979. The Hemeroteca Digital has an important digitalized collection of newspapers, magazines and periodicals, photocopied and scanned, and the optical character recognition technology allows keyword searches, opening up enormous possibilities for researching the careers of musicians unknown to the general public. The aim of this paper is to document the results obtained from this source, describing and mapping Esmeraldino’s main professional activities as a musician: 1) performing on Rádio Tupi and television; 2) accompanying singers in São Paulo’s nightlife, in bars, nightclubs and concert halls; 3) photographic records; and 4) recording phonograms and albums. Through this description and mapping of data obtained from the Hemeroteca Digital, we can get an idea of the repercussions of Esmeraldino Salles’ work in his heyday, as well as the public’s perception of him.

HTR-VT: Handwritten text recognition with vision transformer

We explore the application of Vision Transformer (ViT) for handwritten text recognition. The limited availability of labeled data in this domain poses challenges for achieving high performance solely relying on ViT. Previous transformer-based models required external data or extensive pre-training on large datasets to excel. To address this limitation, we introduce a data-efficient ViT method that uses only the encoder of the standard transformer. We find that incorporate a Convolutional Neural Network (CNN) for feature extraction instead of the original patch embedding and employ Sharpness-Aware Minimization (SAM) optimizer to ensure that the model can converge towards flatter minima yield notable enhancements. Furthermore, our introduction of the span mask technique, which masks interconnected features in the feature map, acts as an effective regularizer. Empirically, our approach competes favorably with traditional CNN-based models on small datasets like IAM and READ2016. Additionally, it establishes a new benchmark on the LAM dataset, currently the largest dataset with 19,830 training text lines. The code will be publicly available at: https://github.com/YutingLi0606/HTR-VT.

Recognizing Digital Ink Chinese Characters Written by International Students Using a Residual Network with 1-Dimensional Dilated Convolution

Due to the complex nature of Chinese characters, junior international students often encounter writing problems related to strokes, components, and their combinations when writing Chinese characters. Digital ink Chinese characters (DICCs) are obtained by sampling the writing trajectory of Chinese characters with a pen input device. DICCs contain rich information, such as the time and space of strokes and sampling points. Recognizing DICCs is crucial for evaluating and correcting writing errors and enhancing the quality of Chinese character teaching for international students. Here, the paper first employs a one-dimensional dilated convolution to digital ink Chinese character recognition (DICCR) and proposes a novel residual network with one-dimensional dilated convolution (1-D ResNetDC). The 1-D ResNetDC not only utilizes multi-scale convolution kernels, but also employs different dilation rates on a single-scale convolution kernel to obtain information from various ranges. Additionally, residual connections facilitate the training of deep one-dimensional convolutional neural networks. Moreover, the paper proposes a more expressive ten-dimensional feature representation that includes spatial, temporal, and writing direction information for each sampling point, thereby improving classification accuracy. Because the DICC dataset of international students is small and unbalanced, the 1-D ResNetDC is pre-trained on the published available dataset. The experiments demonstrate that our approach is effective and superior. This model features a compact architecture, a reduced number of parameters, and excellent scalability.

Digital Discourse on the ChatGPT Controversy: Reflections on the Controversial Use of Artificial Intelligence Among Indonesian Youth

At the end of 2022, a nonprofit technology research institution funded by Altman and Musk released an AI-based chatbot, ChatGPT, which within just three months has shown utility across various industries, particularly in jobs like copywriting, news report writing, customer service, and legal document creation. Its ability to provide coherent and insightful answers, as well as serve as a brainstorming partner, has some college professors concerned that this machine may replace various human jobs (Stokel-Walker, 2022). One likely negative impact is students using this AI-based writing tool to complete academic assignments in the form of essays (Hutson, 2022). Another implication is that researchers (both students and lecturers) may be able to compose scientific texts, partially if not wholly, and escape the radar of AI-written text detection tools (Kim, 2022), as well as peer reviewers (Else, 2023). The methodology used is digital discourse, aiming to reveal how youth discourse influences the presence of the ChatGPT application and how the consumption practice of the application impacts their creativity and critical thinking skills. The findings of this study are expected to contribute to understanding the impact of AI-based chatbots on critical thinking abilities among young people as a basis for development strategies to enhance critical and creative thinking skills in the era of Artificial Intelligence.

Turning a CLIP Model Into a Scene Text Spotter.

We exploit the potential of the large-scale Contrastive Language-Image Pretraining (CLIP) model to enhance scene text detection and spotting tasks, transforming it into a robust backbone, FastTCM-CR50. This backbone utilizes visual prompt learning and cross-attention in CLIP to extract image and text-based prior knowledge. Using predefined and learnable prompts, FastTCM-CR50 introduces an instance-language matching process to enhance the synergy between image and text embeddings, thereby refining text regions. Our Bimodal Similarity Matching (BSM) module facilitates dynamic language prompt generation, enabling offline computations and improving performance. FastTCM-CR50 offers several advantages: 1) It can enhance existing text detectors and spotters, improving performance by an average of 1.6% and 1.5%, respectively. 2) It outperforms the previous TCM-CR50 backbone, yielding an average improvement of 0.2% and 0.55% in text detection and spotting tasks, along with a 47.1% increase in inference speed. 3) It showcases robust few-shot training capabilities. Utilizing only 10% of the supervised data, FastTCM-CR50 improves performance by an average of 26.5% and 4.7% for text detection and spotting tasks, respectively. 4) It consistently enhances performance on out-of-distribution text detection and spotting datasets, particularly the NightTime-ArT subset from ICDAR2019-ArT and the DOTA dataset for oriented object detection.

AI-Integrated Smart Glasses for Enhancing Reading and Guidance Independence for the Visually Impaired

Since there are blind and partially sighted people all over the world, they face significant challenges when it comes to reading. In the past, Braille was the only solution for the blind and visually impaired to read and learn. However, there is a lack of training in Braille and a shortage of Braille machines and facilities. To address these issues, a smart reader for the blind integrated with Raspberry Pi technology has been proposed. The system utilizes YOLO (You only look once) for object detection, OCR (Optical Character Reader) for text recognition, and Google Text-to-Speech to capture and read text aloud through a speaker or headphone. Smart glasses offer real-time assistance and enhance the daily lives of visually impaired individuals. Through a combination of sensors, cameras, and audio feedback systems, these smart glasses enhance the reading capability of the visual impaired. Moreover, advancements in artificial intelligence and Deep learning algorithms further enhance the capabilities of these devices, enabling more accurate personalized assistance.