Handwritten Character Recognition Task Research Articles

Hand Written Character Recognition Using CNN Model

Abstract: Handwritten character recognition (HWR) plays a critical role in document processing, robotic automation, and historical document analysis. While traditional methods relied on template matching and feature engineering, these often struggled with diverse writing styles and noise. Convolutional neural networks (CNNs) have emerged as a powerful alternative, achieving remarkable accuracy in HWR tasks. This paper delves into enhancing HWR performance using CNNs on the EMNIST Balanced dataset. We investigate the impact of data augmentation techniques on model generalizability and propose a custom architecture optimized for the task. Through detailed analysis of performance metrics, confusion matrices, and visualization of predictions, we gain valuable insights into the model's behavior and potential areas for improvement.

A Novel Siamese Network for Few/Zero-Shot Handwritten Character Recognition Tasks

Deep metric learning is one of the recommended methods for the challenge of supporting few/zero-shot learning by deep networks. It depends on building a Siamese architecture of two homogeneous Convolutional Neural Networks (CNNs) for learning a distance function that can map input data from the input space to the feature space. Instead of determining the class of each sample, the Siamese architecture deals with the existence of a few training samples by deciding if the samples share the same class identity or not. The traditional structure for the Siamese architecture was built by forming two CNNs from scratch with randomly initialized weights and trained by binary cross-entropy loss. Building two CNNs from scratch is a trial and error and time-consuming phase. In addition, training with binary cross-entropy loss sometimes leads to poor margins. In this paper, a novel Siamese network is proposed and applied to few/zero-shot Handwritten Character Recognition (HCR) tasks. The novelties of the proposed network are in. 1) Utilizing transfer learning and using the pre-trained AlexNet as a feature extractor in the Siamese architecture. Fine-tuning a pre-trained network is typically faster and easier than building from scratch. 2) Training the Siamese architecture with contrastive loss instead of the binary cross-entropy. Contrastive loss helps the network to learn a nonlinear mapping function that enables it to map the extracted features in the vector space with an optimal way. The proposed network is evaluated on the challenging Chars74K datasets by conducting two experiments. One is for testing the proposed network in few-shot learning while the other is for testing it in zero-shot learning. The recognition accuracy of the proposed network reaches to 85.6% and 82% in few- and zero-shot learning respectively. In addition, a comparison between the performance of the proposed Siamese network and the traditional Siamese CNNs is conducted. The comparison results show that the proposed network achieves higher recognition results in less time. The proposed network reduces the training time from days to hours in both experiments.

A conditional GAN-based approach for enhancing transfer learning performance in few-shot HCR tasks

Supervised learning with the restriction of a few existing training samples is called Few-Shot Learning. FSL is a subarea that puts deep learning performance in a gap, as building robust deep networks requires big training data. Using transfer learning in FSL tasks is an acceptable way to avoid the challenge of building new deep models from scratch. Transfer learning methodology considers borrowing the architecture and parameters of a previously trained model on a large-scale dataset and fine-tuning it for low-data target tasks. But practically, fine-tuning pretrained models in target FSL tasks suffers from overfitting. The few existing samples are not enough to correctly adjust the pretrained model’s parameters to provide the best fit for the target task. In this study, we consider mitigating the overfitting problem when applying transfer learning in few-shot Handwritten Character Recognition (HCR) tasks. A data augmentation approach based on Conditional Generative Adversarial Networks is introduced. CGAN is a generative model that can create artificial instances that appear more real and indistinguishable from the original samples. CGAN helps generate extra samples that hold the possible variations of human handwriting instead of applying traditional image transformations. These transformations are low-level, data-independent operations, and only produce augmented samples with limited diversity. The introduced approach was evaluated in fine-tuning the three pretrained models: AlexNet, VGG-16, and GoogleNet. The results show that the samples generated by CGAN can enhance transfer learning performance in few-shot HCR tasks. This is by achieving model fine-tuning with fewer epochs and by increasing the model’s F1-score and decreasing the Generalization Error (E_{test}).

Backpropagation method modification using Taylor series to improve accuracy of offline neural network training

The aim of this research is to derive new, more effective equations to adjust parameters of a neural network, which have advantage over formulas of the backpropagation method. For this purpose, an objective function of the neural network training is proposed to be modified using expansion of the abstract ideal output of each network layer into a Taylor series. Such decomposition is to be implemented at the point, which is close to the current output of the layer under consideration. The modified objective function gradients are calculated with respect to the difference between ideal and real values of the network parameters. As a result, the obtained weights and biases adjustment equations (for each layer) depend not only on the output of the previous layer, but also on its derivatives. Comparing to the backpropagation method, they allow to reduce the offline neural network training time. The derived formulas can be reduced to the ones of the backpropagation method. This proves the correctness of their development. The efficiency of the new training equations is shown using MNIST handwritten character recognition task. Considering the multilayer structures with different activation functions and number of layers, as far as new formulas are concerned, the same recognition accuracy is achieved, but faster comparing to the conventional backpropagation.

LCSegNet: An Efficient Semantic Segmentation Network for Large-Scale Complex Chinese Character Recognition

Complex scene character recognition is a challenging yet important task in machine learning, especially for languages with large character sets, such as Chinese, which is composed of hieroglyphics with large-scale categories and similar glyphs. Recently, state-of-the-art methods based on semantic segmentation have achieved great success in scene parsing and have been applied in scene text recognition. However, because of limitations in terms of memory and computation, they are only applied in the small category recognition tasks, such as tasks involving English alphabets and digits. In this paper, we propose an efficient semantic segmentation model based on label coding (LC), called LCSegNet, to recognize large-scale Chinese characters. First, to reduce the number of labels, we design a new label coding method based on the Wubi Chinese characters code, called Wubi-CRF. In this method, glyphs and structure information of Chinese characters are encoded into 140-bit labels. Second, we employ an efficient semantic segmentation model for pixel-wise prediction and utilize a conditional random field (CRF) module to learn the constraint rules of Wubi-like coding. Finally, experiments are conducted on three benchmarks: a large Chinese text dataset in the wild (CTW), ICDAR2019-ReCTS, and HIT-OR3C dataset. Results show that the proposed method achieves state-of-the-art performances in both complex scene and handwritten character recognition tasks.

Segmentation and recognition of characters on Tulu palm leaf manuscripts

This paper proposes an efficient method for segmentation and recognition of handwritten characters from Tulu palm leaf manuscript images. The proposed method uses an automated tool with a combination of thresholding and edge detection technique to binarise the image. Further projection profile with connected component analysis is used to line and character segmentation. Deep convolution neural network (DCNN) model used here to extract features and recognise segmented Tulu characters efficiently with a recognition rate of 79.92%. The results are verified using benchmark dataset, the AMADI_LontarSet to generalise our model to handwritten character recognition task. The results showed that our method outperforms from the existing state of art models.

Segmentation and recognition of characters on Tulu palm leaf manuscripts

This paper proposes an efficient method for segmentation and recognition of handwritten characters from Tulu palm leaf manuscript images. The proposed method uses an automated tool with a combination of thresholding and edge detection technique to binarise the image. Further projection profile with connected component analysis is used to line and character segmentation. Deep convolution neural network (DCNN) model used here to extract features and recognise segmented Tulu characters efficiently with a recognition rate of 79.92%. The results are verified using benchmark dataset, the AMADI_LontarSet to generalise our model to handwritten character recognition task. The results showed that our method outperforms from the existing state of art models.

Convolutional ensembles for Arabic Handwritten Character and Digit Recognition.

A learning algorithm is proposed for the task of Arabic Handwritten Character and Digit recognition. The architecture consists on an ensemble of different Convolutional Neural Networks. The proposed training algorithm uses a combination of adaptive gradient descent on the first epochs and regular stochastic gradient descent in the last epochs, to facilitate convergence. Different validation strategies are tested, namely Monte Carlo Cross-Validation and K-fold Cross Validation. Hyper-parameter tuning was done by using the MADbase digits dataset. State of the art validation and testing classification accuracies were achieved, with average values of 99.74% and 99.47% respectively. The same algorithm was then trained and tested with the AHCD character dataset, also yielding state of the art validation and testing classification accuracies: 98.60% and 98.42% respectively.

Handwritten Chinese character recognition using fuzzy image alignment

The task of handwritten Chinese character recognition is one of the most challenging areas of human handwriting classification. The main reason for this is related to the writing system itself which encompasses thousands of characters, coupled with high levels of diversity in personal writing styles and attributes. Much of the existing work for both online and off-line handwritten Chinese character recognition has focused on methods which employ feature extraction and segmentation steps. The preprocessed data from these steps form the basis for the subsequent classification and recognition phases. This paper proposes an approach for handwritten Chinese character recognition and classification using only an image alignment technique and does not require the aforementioned steps. Rather than extracting features from the image, which often means building models from very large training data, the proposed method instead uses the mean image transformations as a basis for model building. The use of an image-only model means that no subjective tuning of the feature extraction is required. In addition by employing a fuzzy-entropy-based metric, the work also entails improved ability to model different types of uncertainty. The classifier is a simple distance-based nearest neighbour classification system based on template matching. The approach is applied to a publicly available real-world database of handwritten Chinese characters and demonstrates that it can achieve high classification accuracy and is robust in the presence of noise.

An irrelevant variability normalization approach to discriminative training of multi-prototype based classifiers and its applications for online handwritten Chinese character recognition

This paper presents an irrelevant variability normalization (IVN) approach to jointly discriminative training of feature transforms and multi-prototype based classifier for recognition of online handwritten Chinese characters. A sample separation margin based minimum classification error criterion is adopted in IVN-based training, while an Rprop algorithm is used for optimizing the objective function. For the IVN approach based on piecewise linear transforms, the corresponding recognizer can be made both compact and efficient by using a two-level fast-match tree whose internal nodes coincide with the labels of feature transforms. Furthermore, the IVN system using weighted sum of linear transforms outperforms that based on piecewise linear transforms. The effectiveness of the proposed approach is first confirmed using an in-house developed online Chinese handwriting corpus with a vocabulary of 9306 characters, and then further verified on a standard benchmark database for an online handwritten character recognition task with a vocabulary of 3755 characters.

Cross-task code reuse in genetic programming applied to visual learning

Abstract We propose a method that enables effective code reuse between evolutionary runs that solve a set of related visual learning tasks. We start with introducing a visual learning approach that uses genetic programming individuals to recognize objects. The process of recognition is generative, i.e., requires the learner to restore the shape of the processed object. This method is extended with a code reuse mechanism by introducing a crossbreeding operator that allows importing the genetic material from other evolutionary runs. In the experimental part, we compare the performance of the extended approach to the basic method on a real-world task of handwritten character recognition, and conclude that code reuse leads to better results in terms of fitness and recognition accuracy. Detailed analysis of the crossbred genetic material shows also that code reuse is most profitable when the recognized objects exhibit visual similarity.

Building compact recognizers of handwritten Chinese characters using precision constrained Gaussian model, minimum classification error training and parameter compression

In our previous work, a so-called precision constrained Gaussian model (PCGM) was proposed for character modeling to design compact recognizers of handwritten Chinese characters. A maximum likelihood training procedure was developed to estimate model parameters from training data. In this paper, we extend the above-mentioned work by using minimum classification error (MCE) training to improve recognition accuracy and using both split vector quantization and scalar quantization techniques to further compress model parameters. Experimental results on a handwritten character recognition task with a vocabulary of 2,965 Kanji characters demonstrate that MCE-trained and compressed PCGM-based classifiers can achieve much higher recognition accuracies than their counterparts based on traditional modified quadratic discriminant function (MQDF) when the footprint of the classifiers has to be made very small, e.g., less than 2 MB.

Learning state machine-based string edit kernels

During the past few years, several works have been done to derive string kernels from probability distributions. For instance, the Fisher kernel uses a generative model M (e.g. a hidden Markov model) and compares two strings according to how they are generated by M. On the other hand, the marginalized kernels allow the computation of the joint similarity between two instances by summing conditional probabilities. In this paper, we adapt this approach to edit distance-based conditional distributions and we present a way to learn a new string edit kernel. We show that the practical computation of such a kernel between two strings x and x ′ built from an alphabet Σ requires (i) to learn edit probabilities in the form of the parameters of a stochastic state machine and (ii) to calculate an infinite sum over Σ * by resorting to the intersection of probabilistic automata as done for rational kernels. We show on a handwritten character recognition task that our new kernel outperforms not only the state of the art string kernels and string edit kernels but also the standard edit distance used by a neighborhood-based classifier.

Invariant pattern recognition using analog recurrent associative memories

A novel invariant pattern recognition approach is proposed based on a special gradient-type recurrent analog associative memory. The system exhibits stable equilibrium points in predefined positions specified by feature vectors extracted from the training set, while invariance to geometrical transformations is inferred by using the tangent distance. Experimental results for handwritten character recognition and face recognition tasks indicate that the proposed approach may yield superior performances over classical solutions based on the Euclidean distance metric. Possible extensions towards modular and sequential pattern recognition are finally outlined.

Writer adaptation for online handwriting recognition

Writer-adaptation is the process of converting a writer-independent handwriting recognition system into a writer-dependent system. It can greatly increasing recognition accuracy, given adequate writer models. The limited amount of data a writer provides during training constrains the models' complexity. We show how appropriate use of writer-independent models is important for the adaptation. Our approach uses writer-independent writing style models (lexemes) to identify the styles present in a particular writer's training data. These models are then updated using the writer's data. Lexemes in the writer's data for which an inadequate number of training examples is available are replaced with the writer-independent models. We demonstrate the feasibility of this approach on both isolated handwritten character recognition and unconstrained word recognition tasks. Our results show an average reduction in error rate of 16.3 percent for lowercase characters as compared against representing each of the writer's character classes with a single model. In addition, an average error rate reduction of 9.2 percent is shown on handwritten words using only a small amount of data for adaptation.

A complete system for NN classification based on a VLSI array processor

This paper describes a VLSI array processor system designed and built for classification problems based on the k-nearest-neighbors approach. This architecture is suitable for different pattern recognition applications and is very efficient for high-dimensional databases. The architecture is scalable with the size of the recognition problem making the system effectively applicable to computational intensive application like on-line pattern recognition. A system prototype composed of a board with two processors, the software driver and a test application have been built and evaluated. For handwritten character recognition task the complete system shows a speed up of 260 times over a sequential algorithm running on a Sun SPARC20 workstation.

Comparison of fast nearest neighbour classifiers for handwritten character recognition

Recently some fast methods ( LAESA and TLAESA) have been proposed to find nearest neighbours in metric spaces. The average number of distances computed by these algorithms does not depend on the number of prototypes and they show linear space complexity. These results where obtained through vast experimentation using only artificial data. In this paper, we corroborate this behaviour when applied to handwritten character recognition tasks. Moreover, we compare LAESA and TLAESA with some classical algorithms also working in metric spaces.