The Support Vector Machines (SVM) classifier is a margin-based supervised machine learning method used for categorization and classification tasks. A Linear SVM classifier uses linear Kernels while a non-linear SVM classifier adopts non-linear Kernels. The Linear SVM classifier is considered an efficient technique, especially for big datasets of high dimensionality in various applications, such as document categorization, time-series classification, outliers’ detection, to name a few. It is much faster to train a linear SVM classifier than a non-linear SVM classifier. For large-scale datasets with various shapes, configurations, and distributions, the computational complexity of training a non-linear SVM classifier is continuously evolving. Current research methods have introduced various problem formulations, solvers, and strategies for speeding up the training process of a non-linear SVM classifier. However, solving a quadratic programming (QP) problem is still challenging, especially for “Big Data”, which poses a great challenge for traditional methods to train a non-linear classifier efficiently. In this paper, we are proposing a novel boosting algorithm to enhance the performance of weak non-linear SVM classifiers using the notion of incremental learning and decremental unlearning. Experimental results over artificial and real datasets with different sizes, shapes, and distributions show that the proposed ensemble boosting algorithm outperforms the individual SVM classifiers measured by the classification accuracy and the Speedup.

A boosted SVM classifier trained by incremental learning and decremental unlearning approach

Handwritten character recognition has been profoundly studied for many years in the field of pattern recognition. Due to its vast practical applications and financial implications, handwritten character recognition is still an important research area. In this research, the Handwritten Ethiopian Character Recognition (HECR) dataset has been prepared to train the model. The images in the HECR dataset were organized with more than one color pen RGB main spaces that have been size normalized to 28 × 28 pixels. The dataset is a combination of scripts (Fidel in Ethiopia), numerical representations, punctuations, tonal symbols, combining symbols, and special characters. These scripts have been used to write ancient histories, science, and arts of Ethiopia and Eritrea. In this study, a hybrid model of two super classifiers: Convolutional Neural Network (CNN) and eXtreme Gradient Boosting (XGBoost) is proposed for classification. In this integrated model, CNN works as a trainable automatic feature extractor from the raw images and XGBoost takes the extracted features as an input for recognition and classification. The output error rates of the hybrid model and CNN with a fully connected layer are compared. A 0.4630 and 0.1612 error rates are achieved in classifying the handwritten testing dataset images, respectively. Thus XGBoost as a classifier performs a better result than the traditional fully connected layer.

/pdf/a-new-hybrid-convolutional-neural-network-and-extreme-3vnuhsb7ww.pdf

A New Hybrid Convolutional Neural Network and eXtreme Gradient Boosting Classifier for Recognizing Handwritten Ethiopian Characters

Handwritten word recognition is undoubtedly a challenging task due to various writing styles of individuals. So, lots of efforts are put to recognize handwritten words using efficient classifiers based on extracted features that rely on the visual appearance of the handwritten text. Due to numerous real-time applications, handwritten word recognition is an important research area which is seeking a lot of attention from researchers for the last 10 years. In this article, the authors have proposed a holistic approach and eXtreme Gradient Boosting (XGBoost) technique to recognize offline handwritten Gurumukhi words. In this direction, four state-of-the-art features like zoning, diagonal, intersection & open-end points and peak extent features have been considered to extract discriminant features from the handwritten word digital images. The proposed approach is evaluated on a public benchmark dataset of Gurumukhi script that comprises 40,000 samples of handwritten words. Based on extracted features, the words are classified into one of the 100 classes based on XGBoost technique. Effectiveness of the system is assessed based on several evaluation parameters like CPU elapsed time, accuracy, precision, recall, F1-score and area under curve (AUC). XGBoost technique attained the best results of accuracy (91.66%), recall (91.66%), precision (91.39%), F1-score (91.14%) and AUC (95.66%) using zoning features based on 90% data as the training set and remaining 10% data as the testing set. The comparison of the proposed approach with the existing approaches has also been done which reveals the significance of the XGBoost technique comparatively.

Offline handwritten Gurumukhi word recognition using eXtreme Gradient Boosting methodology

This work presents a recognition system for Offline Yoruba characters recognition using Freeman chain code and K-Nearest Neighbor (KNN). Most of the Latin word recognition and character recognition have used k-nearest neighbor classifier and other classification algorithms . Research tends to explore the same recognition capability on Yoruba characters recognition . Data were collected from adult indigenous writers and the scanned images were subjected to some level of preprocessing to enhance the quality of the digitized images. Freeman chain code was used to extract the features of THE digitized images and KNN was used to classify the characters based on feature space. The performance of the KNN was compared with other classification algorithms that used Support Vector Machine (SVM) and Bayes classifier for recognition of Yoruba characters. It was observed that the recognition accuracy of the KNN classification algorithm and the Freeman chain code is 87.7%, which outperformed other classifiers used on Yoruba characters.

/pdf/yoruba-handwritten-character-recognition-using-freeman-chain-46aeotw240.pdf

Yoruba Handwritten Character Recognition using Freeman Chain Code and K-Nearest Neighbor Classifier

Medical imaging is an approach used to create visual depictions of body organs. Clinical investigation is required to look at the interior of the body. Therefore to diagnose or view different organs, different image modalities need to be visualized by using certain computational techniques. Computational techniques for biomedical images are becoming a main approach in the area of medical science for detection and diagnosis of diseases. These computational techniques for biomedical images are used by computers to analyze any medical image and to make decisions. In the last few decades, different types of computational techniques have been introduced, and have played a very important role in various areas. These computational techniques using biomedical images are very useful in medical science for the detection of diseases in human body organs. This chapter summarizes several types of imaging modalities like magnetic resonance imaging, medical radiation, ultrasound, fundus, elastography as well as computational techniques such as machine learning, deep learning, and fuzzy techniques to interpret the same. An overview of the mentioned imaging modalities and computational techniques in various image applications such as preprocessing, segmentation, classification, compression, and security have been given. Moreover, challenges and issues that occur in this domain are also presented. A general understanding of the biomedical image modalities and computational techniques for processing these images will help researchers in analyzing biomedical images.

Computational techniques in biomedical image analysis: overview

Selection of classifiers and feature extraction methods has a prime role in achieving best possible classification accuracy in character recognition system. Issues of character recognition system related to choice of classifiers and feature extraction methods can be resolved through these objectives. In this proposed work an efficient Support Vector Machine based off-line handwritten character recognition system has been developed. The experiments have been performed using well known standard database acquired from CEDAR, also seven different approaches of feature extraction techniques have been proposed to construct the final feature vector. It is evident from the experimental results that the performance of Support Vector Machine outperforms other state of art techniques reported in literature.

/pdf/off-line-handwritten-character-recognition-system-using-3etv2yjs4e.pdf

Off-Line Handwritten Character Recognition System Using Support Vector Machine

Removing noise from images is an integral part of the image processing field. Noises can occur in images during acquisition on transmission. The presence of noise can hinder the proper utilization of these images for various applications such as medical imaging, satellite imaging. Use of soft computing methods, especially neural networks, is studied in this paper. Soft computing techniques are preferred for their ability to deal with inexact data and model accurately. Neural network is an approach that uses layers of neurons to arrive at a result. It is modeled after the human brain cell structure. The works done in the field of image denoising using neural networks over the past decade are presented here.

Denoising of Images Using Neural Network: A Review

Image denoising is an important step in the field of image processing. Presence of noise can lead to various obstacles in the way of proper analysis of images to extract information from it like misinterpretation of data, loss in the usability of the image etc. Denoised images are used in various applications such as in medical diagnosis, ultrasound imaging, satellite imaging, pattern recognition etc. Different image denoising techniques are already in existence that uses different filters to remove noise. Fuzzy logic is a soft computing technique that allows for approximations and partial truths. The benefit of using fuzzy logic for denoising purpose is to increase the tractability, robustness and effectiveness of the existing traditional denoising methods. This paper presents a novel fuzzy based method for removal of speckle noise, which mostly affects ultrasound and SAR images.

Region based speckle noise reduction approach using fuzzy techniques

In this work, we present a learning based method focusing on the convolutional neural network (CNN) architecture to detect these forgeries. We consider the detection of both copy-move forgeries and inpainting based forgeries. For these, we synthesize our own large dataset. In addition to classification, the focus is also on interpretability of the forgery detection. As the CNN classification yields the image-level label, it is important to understand if forged region has indeed contributed to the classification. For this purpose, we demonstrate using the Grad-CAM heatmap, that in various correctly classified examples, that the forged region is indeed the region contributing to the classification. Interestingly, this is also applicable for small forged regions, as is depicted in our results. Such an analysis can also help in establishing the reliability of the classification.

Image Forgery Detection with Interpretability

Biometrics is a study of physiological and behavioral characteristics of a person for the purpose of verification. There is an increase in number of biometric traits involved in biometric authentication system depending on the application and the demand in the market. Each modality has its own advantages and disadvantages, and no method is able to satisfy the security needs in completeness. Therefore, there is a continuous effort to look for novel and emerging modalities that are able to provide high security, robust, and also cost-effective. This paper presents some of the novel biometric modalities.

Ankita Katiyar

Papers

Off-Line Handwritten Character Recognition System Using Support Vector Machine

Denoising of Images Using Neural Network: A Review

Region based speckle noise reduction approach using fuzzy techniques

Image Forgery Detection with Interpretability

Novel Biometric Modalities