Input Character Research Articles

Penerapan Jaringan Syaraf Tiruan untuk Pengenalan Pola Huruf Menggunakan Metode Bidirectional Associative Memory (BAM)

The process in artificial intelligence is similar to the process that occurs in the human brain. The pattern recognition process is to determine whether the input and output values ​​are the same so that the system can detect the pattern. A reciprocal relationship between the input and output levels is made possible by the two interconnected layers of the Bidirectional Associative Memory (BAM) technique. As a result, this technique can serve as an associative memory, meaning that some of the data stored in it can be called upon.. The load signal is transmitted from the input layer X to the output layer Y in this procedure. The weight value will be modified by computing the outcomes between the ranks [1,0]. In this work, the active sigmoid function is employed. Three input characters—the letters S, O, and X—are used in this study's 5x5 order matrix. A pattern [27 4] that matches the target is produced by the target letter S [1,-1], a pattern [27 -3] that does not match the target is produced by the target letter O [1.1], and a pattern [-37 21] that matches the target is produced by the target letter X [-1,1]. Consequently, not all patterns are able to meet the established objectives.

Implementation of Software for Learning Message Transmission Using Morse Code

The article presents the implementation of software for training message transmission using Morse code. This is a method of sign coding when letters of the alphabet, numbers, punctuation marks and other symbols are represented as sequences of short and long signals called dots and dashes. It is intended for transmission over serial communication channels. A unique feature of Morse code is the ability to be encoded and decoded by humans without the use of special terminal devices. The paper shows a software implementation of processing input characters using a telegraphic key, it includes: various mathematical requirements and conditions that allow you adjusting to each user, provided that his input speed will either fall or remain unchanged. However, it is possible to adjust the minimum duration threshold in such a way that the system will slowly adapt to a slow increase in input speed. At the same time, the higher the value of the minimum threshold, the faster the adaptation is; however, it will fall within the limit of the maximum input speed that the device can recognize. During the implementation of the program, a flow chart of message transliteration, signal reception and transmission was developed. The whole program implements the following functions: a simple and intuitive translator, with audio output, a simulator for consolidating the skill of entering Morse code using a key; consolidating the skills of recognition and decoding by ear. To create the final application for training, the Processing toolkit was chosen, since its key parameter when choosing a development tool has a simple writing syntax to increase the speed of writing code, as well as the ability of the environment to work with the communication channel of the microcontroller.

Input panel for handwriting detection based on triboelectric nanogenerator

Handwritten signatures are widely used in our daily lives, which focuses on how to obtain handwritten information comprehensively and effectively. Triboelectric nanogenerators can sensitively sense externally applied trigger signals and can be used to collect user handwriting signals and related features for handwriting input character detection and user identification. In this paper, an intelligent handwriting input panel based on triboelectric nanogenerator is proposed, which, combined with deep learning algorithms, achieves an accuracy of 99.32%, 98.96%, 99.14%, and 99.53% for recognizing handwritten signals in Arabic numerals, English words, Chinese characters, and different users, respectively. It is also possible to use different encoding methods for encrypted communication to transmit signals on the proposed handwriting input panel, demonstrating its potential applications. The results show that the smart handwritten input panel has great potential for personal handwritten signature recognition, privacy information, and human–computer interaction.

Design and Evaluation of Controller-Based Raycasting Methods for Efficient Alphanumeric and Special Character Entry in Virtual Reality.

Alphanumeric and special characters are essential during text entry. Text entry in virtual reality (VR) is usually performed on a virtual Qwerty keyboard to minimize the need to learn new layouts. As such, entering capitals, symbols, and numbers in VR is often a direct migration from a physical/touchscreen Qwerty keyboard-that is, using the mode-switching keys to switch between different types of characters and symbols. However, there are inherent differences between a keyboard in VR and a physical/touchscreen keyboard, and as such, a direct adaptation of mode-switching via switch keys may not be suitable for VR. The high flexibility afforded by VR opens up more possibilities for entering alphanumeric and special characters using the Qwerty layout. In this work, we designed two controller-based raycasting text entry methods for alphanumeric and special characters input (Layer-ButtonSwitch and Key-ButtonSwitch) and compared them with two other methods (Standard Qwerty Keyboard and Layer-PointSwitch) that were derived from physical and soft Qwerty keyboards. We explored the performance and user preference of these four methods via two user studies (one short-term and one prolonged use), where participants were instructed to input text containing alphanumeric and special characters. Our results show that Layer-ButtonSwitch led to the highest statistically significant performance, followed by Key-ButtonSwitch and Standard Qwerty Keyboard, while Layer-PointSwitch had the slowest speed. With continuous practice, participants' performance using Key-ButtonSwitch reached that of Layer-ButtonSwitch. Further, the results show that the key-level layout used in Key-ButtonSwitch led users to parallel mode switching and character input operations because this layout showed all characters on one layer. We distill three recommendations from the results that can help guide the design of text entry techniques for alphanumeric and special characters in VR.

Semantic dependency and local convolution for enhancing naturalness and tone in text-to-speech synthesis

Self-attention-based networks have become increasingly popular due to their exceptional performance in parallel training and global context modeling. However, it may fall short of capturing local dependencies, particularly in datasets with strong local correlations. To address this challenge, we propose a novel method that utilizes semantic dependency to extract linguistic information from the original text. The semantic relationship between nodes serves as prior knowledge to refine the self-attention distribution. Additionally, to better fuse local contextual information, we introduce a one-dimensional convolution neural network to generate the query and value matrices in the self-attention mechanism, taking advantage of the strong correlation between input characters. We apply this variant of the self-attention network to text-to-speech tasks and propose a non-autoregressive neural text-to-speech model. To enhance pronunciation accuracy, we separate tones from phonemes as independent features in model training. Experimental results show that our model yields good performance in speech synthesis. Specifically, the proposed method significantly improves the processing of pause, stress, and intonation in speech.

The Design and Evaluation of CAD Custom Batik User Interface

The popularity of batik reflects consumer demands for purchasing fulfillment. Digital technology can bring several advantages and new opportunities to the custom batik design process. Computer-Aided Design (CAD) is a popular digital tool used. This paper will provide the new User Interface (UI) design of CAD custom design batik, namely Batik 4.0. It’s software developed to provide services, such as pattern customization in various sizes, batik character input system, pricing, and production time estimation utilized by the batik industry in Indonesia that can be directly forwarded to the manufacturing process. Research in UI design for CAD batik has not been studied, and this paper will fill the significant gap and upgrade the usability. This evaluation will test the usability of the test, employing the performance matrix and RTA towards UI that already exists. Improvement was made with a wireframe using human-computer interaction (HCI) and usability testing data. User-centered design (UCD) focuses on the role of the user in the process of system development as the wireframe design method. This research shows UI wireframe design for several types of users. In making the UI wireframe design, a usability evaluation was performed again. The evaluation result shows that the new user prototypes have fewer errors and exceed the usability value compared to the previous one. The new design is more user-friendly and can be used as a reference for the future development and improvement of CAD custom batik.

A New Modeling Method for Multiple-Relay Wireless Power Transfer System Considering Cross-Coupling

The cross-coupling among coils makes it complex to analyze and understand the circuit characters of multiple-relay wireless power transfer (MR-WPT) systems. In order to describe and explain the interaction among cross-coupled coils, this paper proposes a novel method named the transfer coefficient model (TCM), which is derived by the LDU factorization of the system's impedance matrix. This model can describe the interaction among cross-coupled coil loops and the composition of each coil current, with clear physical meanings and recursion forms for programming. As a result, the input and output characters of MR-WPT systems can be obtained and explained explicitly. This paper presents the comprehensive and recursion form of TCM for an arbitrary number of coils, followed by the analysis of MR-WPT systems under various operation modes. A five-coil prototype is built to demonstrate the system characters obtained from the TCM. The experimental results coincide with the theoretical analysis, indicating that the TCM is an effective approach for analyzing MR-WPT systems.

What Visibilising Pedagogical Approaches do the Chinese as a Foreign Language (cfl) Teachers Adopt in the Teaching of Chinese Characters?

Abstract Chinese can be considered as one of the more challenging languages to learn for many non-native learners. Its complicated visual script, the Chinese characters, demands that teachers be proficient in utilizing diverse visual pedagogical approaches, coupled with different digital visual tools, to enhance their teaching effectiveness. Adopting a descriptive qualitative analysis method, interview transcripts from fourteen cfl teachers were examined. There were five online teachers, four university teachers and five pre-service teachers. The findings revealed that: 1) all the teachers use various visual pedagogical approaches, such as pictures, videos, Chinese character digital image files, Chinese character studying applications, and websites, to enhance the effectiveness of Chinese character learning, as well as motivate and engage learners; 2) teachers choose different pedagogical approaches to help students visualise characters as a whole unit or as several separate radical components; 3) the use of Pinyin to input characters on digital devices has been widely taught in class; and 4) online teachers, and some pre-service teachers, have replaced traditional handwriting methods with digital character input methods, whereas university teachers have not made this shift. The ways in which these perspectives relate to the form of Chinese characters and developing teaching practices are discussed.

Comparative Analysis of Artificial Neural Network (ANN) and Wavelet Integrated Artificial Neural Network (W-ANN) Approaches for Rainfall Modeling of Southern Rajasthan, India

This paper addresses the challenge of predicting erratic rainfall in Rajasthan state of India, particularly in southern regions. Reliable rainfall predictions are crucial for water resource management and agriculture planning. The research involved selecting 58 stations across seven districts of southern Rajasthan and identifying the best fit computational neural (ANN) and wavelet integrated computational neural (W-ANN) architectures based on performance metrics. Different combinations of input characters, hidden layer neurons, learning algorithms, and training cycles were tested to determine optimal models. Hybrid models, combining wavelet analysis with ANN, were explored to tackle non-stationary hydrologic signals effectively. Results showed that ANN Model C with ten input layer neurons performed best for 74% of stations, followed by Model B (21% of stations) and Model A (5% of stations). Models with increased input and hidden layer neurons performed better. Among the selected stations, 81% of stations demonstrated improved performance using W-ANN models due to effective signal decomposition and information extraction. The hybrid W-ANN models outperformed simple ANN models for rainfall prediction. Both ANN and W-ANN models accurately forecasted weekly rainfall, as observed in the comparison of actual and forecasted values.

Study for Electrooculography Character Input Based on Dual-Channel CNN of Movement Disorder Patients

Study for Electrooculography Character Input Based on Dual-Channel CNN of Movement Disorder Patients

SideNet: Learning representations from interactive side information for zero-shot Chinese character recognition

Existing methods for zero-shot Chinese character recognition usually exploit a single type of side information such as radicals, glyphs, or strokes to establish a mapping with the input characters for the recognition of unseen categories. However, these approaches have two limitations. Firstly, the mappings are inefficient owing to their complexity. Some existing methods design radical-level mappings using a non-differentiable dictionary-matching strategy, whereas others construct sophisticated embeddings to map seen and unseen characters into a unified latent space. Although the latter approach is straightforward, it lacks a learnable scheme for explicit structure construction. Secondly, the complementarity within multiple types of side information has not been effectively explored. For example, the radicals provide structural knowledge at an abstract level, whereas glyphs offer detailed information on their figurative counterparts. To this end, we propose a new method called SideNet that jointly learns character-level representations assisted by two types of interactive side information: radicals and glyphs. SideNet contains a structural conversion module that extracts radical knowledge via dimensional decomposition, and a spatial conversion module that encodes the radical counting map to produce an interactive outcome between radicals and glyph. Finally, we propose a new classifier that integrates the converted features by a similarity-guided fusion mechanism. To the best of our knowledge, this study represents the first attempt to integrate these two types of side information and explore a joint representation for zero-shot learning. Experiments show that SideNet consistently outperforms existing methods by a significant margin in diverse scenarios, including handwriting, printed art, natural scenes, and ancient Chinese characters, which demonstrates the potential of joint learning with multiple types of side information.

Development and user evaluation for a robust character input support system for natural images of electricity reading slips

AbstractText input, which is essential for a variety of applications, is one of the most difficult operations for older people. So, we propose a prototype to automatically detect and input text from a natural image taken from a user's smartphone. For the evaluation of the system, 232 photo data taken by the actual people were prepared. When all data was processed on the server, the average processing time was less than 10 seconds, and the overall recognition rate was over 94%. For user study, 20 healthy older people aged 65 years or older were collected. And, after conducting a simulation with the conventional typing system, the prototype in this proposal was asked to be used. The time required to complete each work was recorded, and the usability was evaluated using System Usability Scale (SUS).As a result, both the healthy older group and the MCI older group tended to shorten the task execution times and increase their SUS scores by using our application, compared to the conventional method. Also, comparing the SUS score with other Web applications, it is considered that usability is high as a prototype.

Enhancing Voice Cloning Quality through Data Selection and Alignment-Based Metrics

Voice cloning, an emerging field in the speech-processing area, aims to generate synthetic utterances that closely resemble the voices of specific individuals. In this study, we investigated the impact of various techniques on improving the quality of voice cloning, specifically focusing on a low-quality dataset. To contrast our findings, we also used two high-quality corpora for comparative analysis. We conducted exhaustive evaluations of the quality of the gathered corpora in order to select the most-suitable data for the training of a voice-cloning system. Following these measurements, we conducted a series of ablations by removing audio files with a lower signal-to-noise ratio and higher variability in utterance speed from the corpora in order to decrease their heterogeneity. Furthermore, we introduced a novel algorithm that calculates the fraction of aligned input characters by exploiting the attention matrix of the Tacotron 2 text-to-speech system. This algorithm provides a valuable metric for evaluating the alignment quality during the voice-cloning process. We present the results of our experiments, demonstrating that the performed ablations significantly increased the quality of synthesised audio for the challenging low-quality corpus. Notably, our findings indicated that models trained on a 3 h corpus from a pre-trained model exhibit comparable audio quality to models trained from scratch using significantly larger amounts of data.

SeDepTTS: Enhancing the Naturalness via Semantic Dependency and Local Convolution for Text-to-Speech Synthesis

Self-attention-based networks have obtained impressive performance in parallel training and global context modeling. However, it is weak in local dependency capturing, especially for data with strong local correlations such as utterances. Therefore, we will mine linguistic information of the original text based on a semantic dependency and the semantic relationship between nodes is regarded as prior knowledge to revise the distribution of self-attention. On the other hand, given the strong correlation between input characters, we introduce a one-dimensional (1-D) convolution neural network (CNN) producing query(Q) and value(V) in the self-attention mechanism for a better fusion of local contextual information. Then, we migrate this variant of the self-attention networks to speech synthesis tasks and propose a non-autoregressive (NAR) neural Text-to-Speech (TTS): SeDepTTS. Experimental results show that our model yields good performance in speech synthesis. Specifically, the proposed method yields significant improvement for the processing of pause, stress, and intonation in speech.

SSMI: Semantic Similarity and Mutual Information Maximization Based Enhancement for Chinese NER

The Chinese NER task consists of two steps, first determining entity boundaries and then labeling them. Some previous work incorporating related words from pre-trained vocabulary into character-based models has been demonstrated to be effective. However, the number of words that characters can match in the vocabulary is large, and their meanings vary widely. It is unreasonable to concatenate all the matched words into the character's representation without making semantic distinctions. This is because words with different semantics also have distinct vectors by the distributed representation. Moreover, mutual information maximization (MIM) provides a unified way to characterize the correction between different granularity of embeddings, we find it can be used to enhance the features in our task. Consequently, this paper introduces a novel Chinese NER model named SSMI based on semantic similarity and MIM. We first match all the potential word boundaries of the input characters from the pre-trained vocabulary and employ BERT to segment the input sentence to get the segmentation containing these characters. After computing their cosine similarity, we obtain the word boundary with the highest similarity and the word group with similarity score larger than a specific threshold. Then, we concatenate the most relevant word boundaries with character vectors. We further calculate the mutual information maximization of group, character and sentence, respectively. Finally, we feed the result from the above steps to our novel network. The results on four Chinese public NER datasets show that our SSMI achieves state-of-the-art performance.

Handwritten Digits and Optical Characters Recognition

The process of transcribing a language represented in its spatial form of graphical characters into its symbolic representation is called handwriting recognition. Each script has a collection of characters or letters, often known as symbols, that all share the same fundamental shapes. Handwriting analysis aims to correctly identify input characters or images before being analysed by various automated process systems. Recent research in image processing demonstrates the significance of image content retrieval. Optical character recognition (OCR) systems can extract text from photographs and transform that text to ASCII text. OCR is beneficial and essential in many applications, such as information retrieval systems and digital libraries.

A Portable One-Hand Braille Input System Enabled by Dual-Interlocked Tactile Sensors for Communication

This article presents a novel portable one-hand Braille input system (BIS) to assist visually impaired users in entering text on electronic devices. The proposed BIS is designed based on a flexible Braille keyboard (FBK), it mainly consists of five dual-interlocked tactile sensors, a processing circuit, and a software modulus. The fully elastomeric graphene/carbon nanotube/ silicone rubber nanocomposites are synthesized and used as tactile sensing material and polydimethylsiloxane (PDMS) as the structural and encapsulation for the fabrication of sensors. Characterization tests showed the tactile sensor has relatively high force sensing sensitivity of 0.649 N <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$^{-\text{1}}$</tex-math> </inline-formula> at sensing range of 0–5 N and a frequency sensing ability of 0.5–2.0 Hz, which can meet human finger pressing requirements. A specific Braille input method is proposed, and personalized parameters are set for users to fit different pressing habits. The average input speed of 3.37 words per minute (WPM) with an average accuracy of 92.2% for Braille characters input is demonstrated by evaluating text entry performance. The results indicated that our developed tactile sensor enabled portable one-hand BIS can assist visual impaired users with enhanced text entry on electronic devices effectively and privately.

VisPhone: Chinese named entity recognition model enhanced by visual and phonetic features

Many Chinese NER models only focus on lexical and radical information, ignoring the fact that there are also certain rules for the pronunciation of Chinese entities. In this paper, we propose VisPhone, which incorporates Chinese characters’ Phonetic features into Transformer Encoder along with the Lattice and Visual features. We present the common rules for the pronunciation of Chinese entities and explore the most appropriate method to encode it. VisPhone uses two identical cross transformer encoders to fuse the visual and phonetic features of the input characters with the text embedding. A selective fusion module is used to get the final features. We conducted experiments on four well-known Chinese NER benchmark datasets: OntoNotes4.0, MSRA, Resume, and Weibo, with F1 scores of 82.63%, 96.07%, 96.26%, 70.79% respectively, improving the performance by 0.79%, 0.32%, 0.39%, and 3.47%. Our ablation experiments have also demonstrated the effectiveness of VisPhone.

Development and User Evaluation for a Robust Character Input Support System for Natural Images of Electricity Reading Slips

Development and User Evaluation for a Robust Character Input Support System for Natural Images of Electricity Reading Slips

HUTNet: An Efficient Convolutional Neural Network for Handwritten Uchen Tibetan Character Recognition.

Recognition of handwritten Uchen Tibetan characters input has been considered an efficient way of acquiring mass data in the digital era. However, it still faces considerable challenges due to seriously touching letters and various morphological features of identical characters. Thus, deeper neural networks are required to achieve decent recognition accuracy, making an efficient, lightweight model design important to balance the inevitable trade-off between accuracy and latency. To reduce the learnable parameters of the network as much as possible and maintain acceptable accuracy, we introduce an efficient model named HUTNet based on the internal relationship between floating-point operations per second (FLOPs) and Memory Access Cost. The proposed network achieves a ResNet-18-level accuracy of 96.86%, with only a tenth of the parameters. The subsequent pruning and knowledge distillation strategies were applied to further reduce the inference latency of the model. Experiments on the test set (Handwritten Uchen Tibetan Data set by Wang [HUTDW]) containing 562 classes of 42,068 samples show that the compressed model achieves a 96.83% accuracy while maintaining lower FLOPs and fewer parameters. To verify the effectiveness of HUTNet, we tested it on the Chinese Handwriting Data sets Handwriting Database 1.1 (HWDB1.1), in which HUTNet achieved an accuracy of 97.24%, higher than that of ResNet-18 and ResNet-34. In general, we conduct extensive experiments on resource and accuracy trade-offs and show a stronger performance compared with other famous models on HUTDW and HWDB1.1. It also unlocks the critical bottleneck for handwritten Uchen Tibetan recognition on low-power computing devices.