Search

The work reported in this article deals with the ground truth generation scheme for online handwritten Bangla documents at text-line, word, and stroke levels. The aim of the proposed scheme is twofold: firstly, to build a document level database so that future researchers can use the database to do research in this field. Secondly, the ground truth information will help other researchers to evaluate the performance of their algorithms developed for text-line extraction, word extraction, word segmentation, stroke recognition, and word recognition. The reported ground truth generation scheme starts with text-line extraction from the online handwritten Bangla documents, then words extraction from the text-lines, and finally segmentation of those words into basic strokes. After word segmentation, the basic strokes are assigned appropriate class labels by using modified distance-based feature extraction procedure and the MLP ( Multi-layer Perceptron ) classifier. The Unicode for the words are then generated from the sequence of stroke labels. XML files are used to store the stroke, word, and text-line levels ground truth information for the corresponding documents. The proposed system is semi-automatic and each step such as text-line extraction, word extraction, word segmentation, and stroke recognition has been implemented by using different algorithms. Thus, the proposed ground truth generation procedure minimizes huge manual intervention by reducing the number of mouse clicks required to extract text-lines, words from the document, and segment the words into basic strokes. The integrated stroke recognition module also helps to minimize the manual labor needed to assign appropriate stroke labels. The freely available and can be accessed at https://byanjon.herokuapp.com/ .

The rapid spread of coronavirus disease has become an example of the worst disruptive disasters of the century around the globe. To fight against the spread of this virus, clinical image analysis of chest CT (computed tomography) images can play an important role for an accurate diagnostic. In the present work, a bi-modular hybrid model is proposed to detect COVID-19 from the chest CT images. In the first module, we have used a Convolutional Neural Network (CNN) architecture to extract features from the chest CT images. In the second module, we have used a bi-stage feature selection (FS) approach to find out the most relevant features for the prediction of COVID and non-COVID cases from the chest CT images. At the first stage of FS, we have applied a guided FS methodology by employing two filter methods: Mutual Information (MI) and Relief-F, for the initial screening of the features obtained from the CNN model. In the second stage, Dragonfly algorithm (DA) has been used for the further selection of most relevant features. The final feature set has been used for the classification of the COVID-19 and non-COVID chest CT images using the Support Vector Machine (SVM) classifier. The proposed model has been tested on two open-access datasets: SARS-CoV-2 CT images and COVID-CT datasets and the model shows substantial prediction rates of 98.39% and 90.0% on the said datasets respectively. The proposed model has been compared with a few past works for the prediction of COVID-19 cases. The supporting codes are uploaded in the Github link: https://github.com/Soumyajit-Saha/A-Bi-Stage-Feature-Selection-on-Covid-19-Dataset.

Online handwriting recognition (OHR) has gained major research interest not just due to the enormous technological advancement in recent years, but also the easy availability of the various electronic devices. This digital revolution is opening up a new dimension in every passing day to the regional and low resource languages with these languages get noticed by the researchers. In this paper, we have targeted a low resource language, Assamese, which is mainly spoken in the eastern region of India. We have proposed a novel and efficient feature vector for recognition of online handwritten Assamese numeral images. Our feature vector has been conceptualized based on the properties of light rays emerging out from a point source. Here we consider that there are multiple hypothetical light emerging sources in a sample numeral image. The amount of light fenced by the image is quantified and considered as a feature. The idea of using point light source to estimate the shape of online handwritten numerals is completely new and efficient. Impressive recognition accuracy is obtained on application of the feature vector on a standard online handwritten Assamese numeral database and it outnumbers some popular and standard feature descriptors, available in the literature. The source code of this work can be found in the following github link: https://github.com/ghoshsoulib/CTRL-Assamese-Digit-Recognition .

In the present work, we have proposed a novel Bangla word segmentation technique that is based on stroke-level busy zone formation procedure. In an unconstrained domain, people often write text where strokes may be poorly aligned (due to multi-directional skewness) and varied combination of strokes with various types of joining between them are possible while forming the words. Hence, a segmentation approach for stroke extraction is pertinent for any stroke-based word recognition system. The presence of a large volume of symbols set (58 basic symbols with more than 280 compound characters) in Bangla script makes the task more challenging. In the current experiment, our stroke-level segmentation approach effectively handles such type of Bangla words. A sub-zoning scheme within busy zone followed by a modified Down->Up->Down (DUD) concept within these sub-zones has been used to find valid segmentation points. This scheme avoids over and under-segmentation issues caused by either inherent writing pattern or due to writing style variations up to certain extent. The proposed segmentation approach has been tested on 6500 online handwritten Bangla word samples with 98.45% correct segmentation accuracy (compared with manually generated ground truth of the same database).

This paper explores the domain of online handwritten Bangla character recognition by stroke-based approach. The component strokes of a character sample are recognized firstly and then characters are constructed from the recognized strokes. In the current experiment, strokes are recognized by both supervised and unsupervised approaches. To estimate the features, images of all the component strokes are superimposed. A mean structure has been generated from this superimposed image. Euclidian distances between pixel points of a stroke sample and mean stroke structure are considered as features. For unsupervised approach, K-means clustering algorithm has been used whereas six popular classifiers have been used for supervised approach. The proposed feature vector has been evaluated on 10,000-character database and achieved 90.69% and 97.22% stroke recognition accuracy in unsupervised (using K-means clustering) and supervised way (using MLP [multilayer perceptron] classifier). This paper also discusses about merit and demerits of unsupervised and supervised classification approaches.

Made available in DSpace on 2022-04-28T19:52:56Z (GMT). No. of bitstreams: 0 Previous issue date: 2022-06-01 Background and Objective: Cervical cancer is one of the leading causes of women's death. Like any other disease, cervical cancer's early detection and treatment with the best possible medical advice are the paramount steps that should be taken to ensure the minimization of after-effects of contracting this disease. PaP smear images are one the most effective ways to detect the presence of such type of cancer. This article proposes a fuzzy distance-based ensemble approach composed of deep learning models for cervical cancer detection in PaP smear images. Methods: We employ three transfer learning models for this task: Inception V3, MobileNet V2, and Inception ResNet V2, with additional layers to learn data-specific features. To aggregate the outcomes of these models, we propose a novel ensemble method based on the minimization of error values between the observed and the ground-truth. For samples with multiple predictions, we first take three distance measures, i.e., Euclidean, Manhattan (City-Block), and Cosine, for each class from their corresponding best possible solution. We then defuzzify these distance measures using the product rule to calculate the final predictions. Results: In the current experiments, we have achieved 95.30%, 93.92%, and 96.44% respectively when Inception V3, MobileNet V2, and Inception ResNet V2 run individually. After applying the proposed ensemble technique, the performance reaches 96.96% which is higher than the individual models. Conclusion: Experimental outcomes on three publicly available datasets ensure that the proposed model presents competitive results compared to state-of-the-art methods. The proposed approach provides an end-to-end classification technique to detect cervical cancer from PaP smear images. This may help the medical professionals for better treatment of the cervical cancer. Thus increasing the overall efficiency in the whole testing process. The source code of the proposed work can be found in github.com/rishavpramanik/CervicalFuzzyDistanceEnsemble. Department of Computer Science and Engineering, Jadavpur University, 188 Raja S C Mallick Rd, West Bengal Department of Metallurgical and Material Engineering, Jadavpur University, 188 Raja S C Mallick Rd, West Bengal Department of Computer Science and Technology, University of Engineering and Management, West Bengal Department of Computing, São Carlos Federal University-UFScar, São Paulo Regensburg Medical Image Computing (ReMIC), Ostbayerische Technische Hochschule Regensburg (OTH Regensburg), Bavaria Department of Computing, São Paulo State University, Av. Eng. Luiz Edmundo Carrijo Coube, 14-01, São Paulo

The analysis of human facial expressions from the thermal images captured by the Infrared Thermal Imaging (IRTI) cameras has recently gained importance compared to images captured by the standard cameras using light having a wavelength in the visible spectrum. It is because infrared cameras work well in low-light conditions and also infrared spectrum captures thermal distribution that is very useful for building systems like Robot interaction systems, quantifying the cognitive responses from facial expressions, disease control, etc. In this paper, a deep learning model called IRFacExNet (InfraRed Facial Expression Network) has been proposed for facial expression recognition (FER) from infrared images. It utilizes two building blocks namely Residual unit and Transformation unit which extract dominant features from the input images specific to the expressions. The extracted features help to detect the emotion of the subjects in consideration accurately. The Snapshot ensemble technique is adopted with a Cosine annealing learning rate scheduler to improve the overall performance. The performance of the proposed model has been evaluated on a publicly available dataset, namely IRDatabase developed by RWTH Aachen University. The facial expressions present in the dataset are Fear, Anger, Contempt, Disgust, Happy, Neutral, Sad, and Surprise. The proposed model produces 88.43% recognition accuracy, better than some state-of-the-art methods considered here for comparison. Our model provides a robust framework for the detection of accurate expression in the absence of visible light.

Breast cancer is one of the deadliest diseases in women and its incidence is growing at an alarming rate. However, early detection of this disease can be life-saving. The rapid development of deep learning techniques has generated a great deal of interest in the medical imaging field. Researchers around the world are working on developing breast cancer detection methods using medical imaging. In the present work, we have proposed a two-stage model for breast cancer detection using thermographic images. Firstly, features are extracted from images using a deep learning model, called VGG16. To select the optimal subset of features, we use a meta-heuristic algorithm called the Dragonfly Algorithm (DA) in the second step. To improve the performance of the DA, a memory-based version of DA is proposed using the Grunwald-Letnikov (GL) method. The proposed two-stage framework has been evaluated on a publicly available standard dataset called DMR-IR. The proposed model efficiently filters out non-essential features and had 100% diagnostic accuracy on the standard dataset, with 82% fewer features compared to the VGG16 model.

The novel SARS-CoV-2 virus, responsible for the dangerous pneumonia-type disease, COVID-19, has undoubtedly changed the world by killing at least 3,900,000 people as of June 2021 and compromising the health of millions across the globe. Though the vaccination process has started, in developing countries such as India, the process has not been fully developed. Thereby, a diagnosis of COVID-19 can restrict its spreading and level the pestilence curve. As the quickest indicative choice, a computerized identification framework ought to be carried out to hinder COVID-19 from spreading more. Meanwhile, Computed Tomography (CT) imaging reveals that the attributes of these images for COVID-19 infected patients vary from healthy patients with or without other respiratory diseases, such as pneumonia. This study aims to establish an effective COVID-19 prediction model through chest CT images using efficient transfer learning (TL) models. Initially, we used three standard deep learning (DL) models, namely, VGG-16, ResNet50, and Xception, for the prediction of COVID-19. After that, we proposed a mechanism to combine the above-mentioned pre-trained models for the overall improvement of the prediction capability of the system. The proposed model provides 98.79% classification accuracy and a high F1-score of 0.99 on the publicly available SARS-CoV-2 CT dataset. The model proposed in this study is effective for the accurate screening of COVID-19 CT scans and, hence, can be a promising supplementary diagnostic tool for the forefront clinical specialists.

This paper proposes a model which is used to recognize online handwritten Bangla words. This word recognition module comprises of different modules: preprocessing of the word samples, segmentation of words into basic strokes, recognizing the basic strokes using multilayer perceptron, followed by recognition of words using Hidden Markov Model (HMM) aided by Language Model (LM). For stroke recognition, two different feature extraction techniques (point-based and curvature-based procedures) are used using late fusion technique. Top 5 stroke recognition choices are used to construct HMM for the prediction of word sample. An N-gram LM is applied as a post-processing step to rectify the HMM outcomes if required. A total of 50 different word samples with 110 instances each are used to evaluate the proposed model. The overall stroke-level and word-level recognition accuracies obtained by this model are 95.4% and 90.3%, respectively. The proposed model can be extended to recognize online handwritten words written in other script like Devanagari, Assamese, and Gurumukhi, etc. The methodologies described in the manuscript can also be applied for offline word recognition purpose.

Western countries entertain speech recognition-based applications. It does not happen in a similar magnitude in East Asia. Language complexity could potentially be one of the primary reasons behind this lag. Besides, multilingual countries like India need to be considered so that language identification (words and phrases) can be possible through speech signals. Unlike the previous works, in this paper, we propose to use speech signal patterns for spoken language identification, where image-based features are used. The concept is primarily inspired from the fact that speech signal can be read/visualized. In our experiment, we use spectrograms (for image data) and deep learning for spoken language classification. Using the IIIT-H Indic speech database for Indic languages, we achieve the highest accuracy of 99.96%, which outperforms the state-of-the-art reported results. Furthermore, for a relative decrease of 4018.60% in the signal-to-noise ratio, a decrease of only 0.50% in accuracy tells us the fact that our concept is fairly robust.

Feature selection through optimization techniques provides an interesting approach to minimize computational time with enhanced prediction capability, and has better cognizance of data in any pattern recognition application. This paper is an extended version of previously published work (Sen et al. in: 7th IAPR TC3 workshop on artificial neural networks in pattern recognition, Ulm, Germany, pp 246–256, 2016), where a quad-tree based image segmentation approach has been discussed to estimate some topological and shape based features (192-attributed) for the recognition of online handwritten Bangla characters. The previous work achieved a recognition accuracy of 98.5% on a database consisting of 10,000 handwritten Bangla characters. In this paper, parameters, used in the previous version during feature estimation, are tuned to improve the performance of the overall system. Thereafter, krill-herd a bio-inspired, meta-heuristic algorithm has been applied to find the optimal feature vector by reducing the dimension of the original feature vector. The reduced feature vector has been fed to Sequential Minimal Optimization classifier for the recognition of the same online handwritten Bangla character database used in previous work. It has been observed that result obtained with this optimal feature set is almost equivalent as the result produced by the entire feature vector.

The transfer learning approach has eradicated the need for running the Convolutional Neural Network (CNN) models from scratch by using a pre-trained model with pre-set weights and biases for recognition of different complex patterns. Going by the recent trend, in this work, we have explored the transfer learning approach to recognize online handwritten Bangla and Devanagari basic characters. The transfer learning models considered here are VGG-16, ResNet50, and Inception-V3. To impose some external challenges to the models, we have augmented the training datasets by adding different complexities to the input data. We have also trained these three transfer learning models from scratch (i.e., not using pre-set weights of the pre-trained models) for the same recognition tasks. Besides, we have compared the outcomes of both the procedures (i.e., running from scratch and by using pre-trained models). Results obtained by the models are promising, thereby establishing its effectiveness in developing a comprehensive online handwriting recognition system.

The present experiment deals with online handwriting recognition for two regional languages Bangla and Devanagari using CNN (convolution neural network) a deep learning concept. Our proposed model consists of two convolution and pooling layers and a fully connected network. CNN model give relaxation of producing handcrafted features manually rather generates features, reduce feature dimension automatically and pass the features to a fully connected network for classification purpose. The current experiment is performed on 10000 Bangla basic characters and 1800 Devanagari characters. We have achieved 99.65% and 98.87% recognition accuracy for Bangla and Devanagari character datasets respectively.

In spite of the introduction of modern technologies, handwriting is still considered as one of the key means of communication and disseminating information in our daily life. Though commercial handwriting recognizers are available for several Western languages but same is not true for Bangla. The complexity of the script is one of the reasons. Recognizing both online and offline handwriting has pros and cons of their own. In this paper, a deep-learning based approach has been adopted to recognize isolated Bangla characters in both online and offline modes. The experiments have been performed on a database of 15000 handwritten Bangla character samples and highest accuracies of 99.91% and 99.92% have been obtained for offline and online modes respectively.

This article presents a set of novel features for robust online Bangla handwritten character recognition. Two feature extraction methods are presented here. The first describes the transition from background to foreground pixels and vice versa. The second uses a combination of topological features and centre-of-gravity- (CG) based circular features where global information, local information, and Circular Quadrant Mass Distribution information have been extracted. The impact of each along with their combination have also been analyzed. A total of 15,000 isolated online Bangla character samples have been collected and used for the evaluation. A Support Vector Machine classifier records the best recognition rate when the transition count feature, CG-based circular features, and topological features are combined.

This paper inspects the impact of feature vector produced by Frechet Distance (FD) along with the conventional distance based features to recognize online handwritten Bangla characters. FD based feature computation starts with dividing a character sample into different rectangular zones. Then FD values are computed from each zone to every other zones. In distance based feature extraction technique also a character sample is divided into several segments and distances are measured from a particular segment to all other segments. Feature vectors so produced are experimented on 10000 online handwritten Bangla characters. SVM (Support Vector Machine) classifier produces the reasonably satisfactory recognition accuracy of 98.98% when FD based features are combined with distance based features.

Nowadays, researchers aim to enhance man-to-machine interactions by making advancements in several domains. Facial emotion recognition (FER) is one such domain in which researchers have made significant progresses. Features for FER can be extracted using several popular methods. However, there may be some redundant/irrelevant features in feature sets. In order to remove those redundant/irrelevant features that do not have any significant impact on classification process, we propose a feature selection (FS) technique called the supervised filter harmony search algorithm (SFHSA) based on cosine similarity and minimal-redundancy maximal-relevance (mRMR). Cosine similarity aims to remove similar features from feature vectors, whereas mRMR was used to determine the feasibility of the optimal feature subsets using Pearson&rsquo, s correlation coefficient (PCC), which favors the features that have lower correlation values with other features&mdash, as well as higher correlation values with the facial expression classes. The algorithm was evaluated on two benchmark FER datasets, namely the Radboud faces database (RaFD) and the Japanese female facial expression (JAFFE). Five different state-of-the-art feature descriptors including uniform local binary pattern (uLBP), horizontal&ndash, vertical neighborhood local binary pattern (hvnLBP), Gabor filters, histogram of oriented gradients (HOG) and pyramidal HOG (PHOG) were considered for FS. Obtained results signify that our technique effectively optimized the feature vectors and made notable improvements in overall classification accuracy.

In the present experiment, we have investigated the effectiveness of two handcrafted feature extraction techniques for the recognition of constituent strokes of online handwritten Bangla character samples. These techniques estimate local and global shape information from a stroke sample. Combined feature vector is fed to Multi-Layer Perceptron (MLP)-based classifier for stroke recognition purpose. We have achieved 91.27% recognition accuracy over test set. In the current experiment, total size of the stroke database is 32,534. Among the samples, 30% of the entire strokes are considered as test set and rest are used to train the recognition model. Afterward, we have implemented a Deterministic Finite Automata (DFA)-based character recognition system from the recognized strokes. Final outcome of the system is satisfactory considering the stroke variation orders while writing Bangla characters.

In the present work, a typical Convolutional Neural Network (CNN) architecture has been used for the recognition of online handwritten isolated Bangla characters. A detailed analysis about the effects of using different kernel variations, pooling strategies, and activation functions in the CNN architecture has been performed. In this work, total 10000 character samples have been used and among the samples, 30% have been considered as test set and rest 70% have been used to train the recognition model. On test dataset, the technique has been provided 99.40% recognition accuracy. The outcome is better than some recently proposed handcrafted features used for the recognition of online handwritten Bangla characters.

This paper deals with stroke-based online Bangla character recognition strategy. In the present work, constituent strokes have been extracted from characters and then popularly used distance based features have been estimated in order to recognize the basic strokes. Next, a rule-based approach is followed for the recognition of the characters from the previously recognized strokes. A total of 15,000 isolated online handwritten Bangla characters contributing 32,534 stroke samples have been used in this experiment, and a satisfactory result of 89.39% recognition accuracy has been achieved.

In this paper, an effort has been made to emphasize the usefulness of Hausdorff Distance (HD) and Directed Hausdorff Distance (DHD) based features for the recognition of online handwritten Bangla basic characters. Every character sample is divided into N number of rectangular zones and then HD- and DHD-based features have been computed from every zone to every other zone. These distance measurements are served as feature values for the present work. Experiment has been done on a set of 10,000 character dataset. Multilayer Perceptron (MLP) produces the best result with an accuracy of 95.57% when sample character is divided into 16 rectangular zones and DHD-based procedure has been considered.