Search

In this paper, we obtain samples of thermal aging transformer insulation paper at 6 time points through an accelerated thermal aging test that simulates the actual operating state of the transformer. The Raman spectroscopy diagnosis model of transformer insulating paper aging is proposed. The model combines an autoencoder and a linear discriminant analysis algorithm to extract the characteristics of the Raman spectrum, and uses a Naive Bayes classification method to divide the aging stage of the insulating paper. Through the Raman spectroscopy diagnosis model of transformer insulation paper aging designed in this paper, the Raman spectroscopy characteristic quantity extracted by the AE-LDA method can effectively reflect the aging characteristic information of the insulating paper, and the Naive Bayes classification algorithm is used to predict. Research shows that the model has a high predictive accuracy for the assessment of the aging state of insulating paper, and can effectively diagnose the aging state of transformer oil-paper insulation. The Raman spectroscopy used in the diagnosis of the aging of insulating paper has good aging discrimination ability and good application prospects, and provides a new idea for the diagnosis of the aging state of transformer oil-paper insulation.

Because the traditional insulation paper state detection method has some shortcomings, this paper proposes a non-destructive detection method, which uses the texture feature value of insulation paper image to establish a relationship with the degree of polymerization (DP) of insulation paper state, and then indirectly estimates the aging state of insulation paper. Firstly, the insulation paper images in different aging states are collected, and the texture feature values of the images are extracted by gray histogram, gray level co-occurrence matrix and gray level run-length matrix; then, the texture feature values are selected by correlation; then, the validity of the selected texture feature values to represent the insulation paper state is verified by support vector machine (SVM); finally, the multiple linear regression method is used. The relationship between the degree of polymerization and the texture feature value is established and verified by the actual results. The results show that the texture feature value has a good fitting relationship with the degree of polymerization, which verifies the effectiveness of this method in evaluating the state of insulating paper.

With the rapid growth of scientific publications, research paper recommendation which suggests relevant research papers to users can bring great benefits to researchers. In this paper, we focus on the problem of recommending conference papers to the conference attendees. While most of the related existing methods depend on the content-based filtering, we propose a unified conference paper recommendation method named $CPRec$ , which exploits both the contents and the authorship information of the papers. In particular, besides the contents, we exploit the relationships between a user and the authors of a paper for recommendation. In our method, we extract several features for a user-paper pair from the citation network, the coauthor network, and the contents, respectively. In addition, we derive a user’s pairwise preference towards the conference papers from the user’s bookmarked papers in each conference. Furthermore, we employ a pairwise learning to rank model which exploits the pairwise user preference to learn a function that predicts a user’s preference towards a paper based on the extracted features. We conduct a recommendation performance evaluation using real-world data and the experimental results demonstrate the effectiveness of our proposed method.

To improve the accuracy of user implicit rating prediction, we combine the traditional latent factor model (LFM) and bidirectional gated recurrent unit neural network (BiGRU) model to propose a hybrid model that deeply mines the latent semantics in the unstructured content of the text and generates a more accurate rating matrix. First, we utilize the user’s historical behavior (favorites records) to build a user rating matrix and decompose the matrix to obtain the latent factor vectors of users and literature. We also apply the BERT model for word embedding of the research papers to obtain the sequence of word vectors. Then, we apply the BiGRU with the user attention mechanism to mine the research paper textual content and to generate the new literature latent feature vectors that are used to replace the original literature latent factor vectors decomposed from the rating matrix. Finally, a new rating matrix is generated to obtain users’ ratings of noninteractive research papers and to generate the recommendation list according to the user latent factor vector. We design experiments on the real datasets and verify that the research paper recommendation model is superior to traditional recommendation models in terms of precision, recall, F1-value, coverage, popularity and diversity.

In the last several years, we published several papers to address the problem of Face Identification. The techniques employed in those articles were implemented in transform domains. The Discrete Cosine (DCT) and the Discrete Wavelet (DWT) Transforms were utilized, either combined or individually, to extract features which form the final model for each participant in a given dataset. In this paper, we highlight significant parts of our previous works in order to give a fair comparison among all approaches. The results included here are for the following datasets: ORL, YALE, FERET, FEI, Georgia Tech, and Cropped AR. Features are DWT, DCT, energy-based selected DCT-DWT, and combined DCT-DWT coefficients while the classifier is Euclidean distance, either squared or with power of one.

Forgery of paper invoices and certificates negatively causes huge economic loss every year. In the molding process of paper, plant fibres inside will distribute randomly and form the unique microscopic surface. Accordingly, forensic researchers have presented many techniques to build its feature descriptor for paper anti-counterfeiting. Among those methods, normal vector field and reconstructed surface based on multi-view photos captured by mobile camera have indicated incredibly effectiveness and practicality. However, the inefficiency of existing models using mobile camera is the neck in real time detection scenarios. In this paper, we focus on the efficient microstructure orientation estimation of paper surface for authentication. To address this problem, we investigate the reflection characteristics of rough surface and quantify the light attenuation accurately. Then, we model the entire propagation path of light emitted by flash, and exploit single photo captured by mobile camera to reconstruct microscopic inclination based on beam energy loss. Reduction of required pictures significantly enhances forensics efficiency. Through comparison experiments, we demonstrate that our proposed method can be used for paper discrimination effectively with speedy feature extraction.

Paper influence analysis is essential technology of the literature management system. In recent years, research on influence analysis has received crucial attention from the industry and academia. In this field, there are many new challenges in how to evaluate valuable papers in the paper citation network. Therefore, in this paper, we have proposed an influence model in paper citation networks with integration features of PageRank and Hyperlink-Induced Topic Search (HITS). First, three ranking features are extracted with the combination of PageRank value, hub, and authority. Second, more features are added to our ranking classification method. Third, the ranking-based classification with $P$ ageRank and HITS is proposed to analyze the paper propagation. An influence score is computed to evaluate the importance of a paper. The experiments show that our ranking-based influence model is effective for paper propagation.

As the connectivity of autonomous vehicles keeps growing, it is an accepted fact that they are even more vulnerable to malicious cyber-attacks. Recently, sensor spoofing has become an emerging attack that can compromise vehicle safety as vehicles are equipped with more sensors. Thus, it is critical to validate the sensor readings before utilizing them for future actions. In this paper, we develop HDAD, a hyperdimensional computing-based anomaly detection method. Hyperdimensional computing (HDC) is an emerging brain-inspired computing paradigm that mimics the brain cognition and leverages hyperdimensional vectors with fully distributed holographic representation and (pseudo)randomness. The key idea of HDAD is to use HDC to build encoder and decoder to reconstruct the sensor readings. The anomalous data typically have comparatively higher reconstruction errors than normal sensor readings. We explore three different metrics to measure the reconstruction error including mean squared error, mean absolute error, and cosine similarity. Using a real-world vehicle sensor reading dataset, we demonstrate the feasibility and efficacy of HDAD, opening the door for a new set of anomaly detection algorithm design.

Over the last few years, web data extraction has gained popularity. Product information on the Ecommerce website floods the internet with big data. Web-based business sites these days have gotten one of the most significant hotspots for getting a large amount of relevant data. Wide range of software application designs to extract relevant data from web pages in order to draw in more business. The extracted data can be used for retailer business and data analysis purposes. The web pages on such sites are based on different technologies, and the generated web documents are in structured or unstructured formats. Manually extract such relevant product data and multimedia type Information from the websites is complex and time-consuming. After extraction of data needs to be classified because web content contains unwanted data e.g. design information, advertising content. This paper describes different Procedures for web document classification and extraction.

In order to automate the task of correcting handwritten subjective answers, we envision a system that reads such answers and grades them. In this project, we have created a mobile application to achieve this task. We have used the RAKE algorithm to extract key-phrases, which are then embedded using Word2vec model and compared using Word Mover's Distance. Our created model achieved almost human like evaluation accuracy. We have also provided accessibility features to the users with the aim of making this application a one-stop solution to automatic evaluation of handwritten subjective answers.

Biological information can easily be monitored by installing sensors in a lavatory bowl. Lavatories are usually shared by several people, so users need to be identified. Because of the need for privacy, using cameras, microphones, or scales is not appropriate. Though personal identification can be done using a touch panel, the user may forget to use it because the action is not necessary. In this paper, we focus on the differences in the way of pulling a toilet paper roll and propose a system that identifies individuals based on features of rotating of toilet paper rolls with a gyroscope. The evaluation results revealed that 83.9% accuracy was achieved for a five-person group in a laboratory environment, and 69.2% accuracy was achieved for a five-person group in a practical environment.

Nowadays, there is a rapidly increasing number of conferences and journals in computer science that make a lot of challenges for researchers to find an appropriate venue to submit their scientific work. There is a need for a recommendation system that can support researchers for a better process of paper submission. In this paper, we present an efficient approach for building such a recommendation model by using embedding methods, Global Vector (GloVe) 1 created by Pennington et al. [1] and FastText 2 proposed by Facebook [2], Convolutional Neural Network (CNN) [3], and LSTM. The experimental results show that the combination of CNNs and FastText, CNN + FastText, can achieve the best performance in terms of the Top 1 Accuracy compared with other techniques, including the S2RSCS model, as presented in [4]. Moreover, the performance by using GloVe or FastText is much better, faster, and more stable than S2RSCS in most cases.

Now a days due to the development in color printing technology the rate of counterfeit notes production and distribution is increasing. This is a massive problem, faced by almost all the countries. It affects the economy, sine it compromises the security of the real economy. Such counterfeit currencies are used to fuel nefarious motives, usually involving terrorist activities. According to the research, developing countries like India have been impacted by this very negatively. Even after the steps taken in 2016 to remove the counterfeits, by executing the demonetization of 500 and 1000 rupees bank notes in India the counterfeits of the new notes have begun circulating. This is due to the highly advanced technology adopted by the counterfeiters which makes the tracking of these counterfeit notes hard. This has become a very critical issue and the negative impact due to the counterfeit currency keeps rising. The only one solution for this problem for a common man is to detect the fake currency, by using the fake currency detector machine. These machines are used in banks and large scale business, but for a small business or for a common man these machines are not affordable. This paper gives the complete methodology of fake note detector machine, which is affordable even for a common man. By implementing the applications of image processing techniques we can find out whether the currency notes are fake or not. Image processing technique consists of a number of operations that can be performed on an image, some of which include image segmentation, edge detection, gray scale conversion etc. The proposed system will have advantages like simplicity, reliability and costs less.

Oil paper insulation is the main insulation form of large transformers. Partial discharge(PD) in oil paper insulation will not only damage the insulation performance, but also is the precursor and manifestation of insulation deterioration. Weibull distribution parameter is an important characteristic parameter of PD. In this paper a method is proposed to extract the characteristic parameter of PD based on the three-parameter Weibull model, then gray wolf optimization algorithm and particle swarm optimization algorithm are used to calculate the parameters of the model. The result shows that the two algorithms can get the parameters quickly and accurately, and the model can effectively reflect the characteristics of the $H_{n}(q)$ spectrum of PD times.

In recent years, with the increasing investment in scientific research, the number of papers has proliferated, and the development of scientific research has also received widespread attention from scholars. Paper citations as an indicator to measure academic influence have essential reference value for future research directions. Aiming at the future paper citation number prediction problem, we propose a citation number prediction model based on the recurrent neural network method with gated recurrent unit(GRU-CPM). In this paper, we first extract features from real data sets that are useful for predicting the number of citations in papers, and then the features were input into GRU-CPM for prediction. Finally, the prediction results are compared with other regression models. Experimental results demonstrate that the GRU-CPM has higher prediction accuracy and faster convergence speed. The time series prediction of citation count is better than existing methods.

Electrocardiogram (ECG) represents the electrical activity of the heart which is used for the diagnosis purpose by the doctors. This paper proposed a method of converting printed ECG data into digital electrocardiogram (ECG) signals. We have developed a system that converts scanned ECG stripped taken from MIT-BIH Normal Sinus Rhythm/ Arrhythmia Database and also data from real patients’ records taken from hospitals into samples as text using digital image processing techniques. Digital data obtained using our proposed model retains all the ECG information of the raw data with precision. The proposed method is reliable for diagnosing ECG based analysis and different cardiac diseases.

Speech Emotion Recognition (SER) is the technique for determining underlying emotions from speech samples. Usually text transcripts supplement vocal cues and contains additional information that boosts the SER process. An unsupervised Fuzzy Inference System (FIS) is developed for SER, in this paper, that incorporates audio and text features. The extracted features are: pitch, energy and textual sentiment score. The proposed system is based on Mamdani Fuzzy Inference model and is capable of determining four emotions: happy, sad, angry and neutral. Our FIS has three variants based on the sentiment lexicon- AFINN, SentiWordNet and VADER selected for computing textual sentiment score as the text feature. The main highlights of this work are: i) formulation of eleven novel fuzzy rules based on audio and text cues for SER ii) comparative analysis of all variants of our proposed unsupervised FIS with six state-of-the-art supervised machine learning approaches for SER iii) both speaker independent SER and speaker dependent SER are executed iv) investigations reveal that for few speakers in speaker-dependent SER have received higher accuracies than the others and v) the proposed unsupervised FIS can handle multiple datasets without any training while the supervised machine learning algorithms fail for cross-dataset evaluation. The experiments conducted on speech datasets: SAVEE and RAVDESS, indicate that our FIS has achieved higher accuracy and f1-scores as compared to the other state-of-the-art-methods.

At the present time evaluating an exam papers and declaring end result in a restricted period of time is a difficult task for educational schools, colleges, institutions, departments and Universities. Thus manual exam paper correction becomes more difficult and many fraudulent activities are happened today. To make it easier and more accurate the proposed aim is to develop a software for automatic exam paper evaluation and grading system, the system works by scanning the handwritten written exam papers then the scanned image be improved into an editable text using OCR tool and the evaluation will perform by matching the key terms which is maintained in the template. It is entirely integrated approach upon dissimilar level of knowledge by the method of examination, evaluation, result and formulation of subject papers. In a field of education though teaching, evaluation and the performance method many organizations initiated with the use of advanced technologies. The approach of evaluating the examination papers is evolved although use of computer wherever the utilization of computer is obligatory in every varied techniques for analysis. This paper structured in different patterns such as offline, online and the manual exam paper evaluation with different analysis and techniques.

Traditional signature-based malware detection techniques have been used for many years owing to their high detection rates and low false positive rates. However, signature-based techniques are ineffective as they do not detect new, polymorphic or transient malware. To beat weaknesses in signature-detection techniques, researchers have turned to behaviour-based ones, which create a malware behaviour profile by capturing malicious API calls throughout execution. In this context, API matching techniques use API calls to calculate similarities between malware. However, API concatenation techniques need significant API call process resources, which makes these methods slow owing to process quality, and thus cannot scale to large API call sequences outside the lab. In this paper, we review present malware detection strategies supported by API call sequences.

Agriculture plays an important role in farmer’s life. Sometimes manual identification of disease is time consuming and need of labor is more. One of the most important facts that reduce the growth of plants is disease attack. Overall study about agriculture shows that quality and quantity of agricultural products may be reduced due to various factors of plant diseases. These diseases can be more easily identified by using machine learning approach as compared to manual method. Hence machine learning method can be used to identify the affected leaf images.Images captured by camera will be processed using different image processing technique. These techniques will help in identifying plant diseases thereby increasing the yield of plants. This survey paper describes plant disease identification using Machine Learning Approach and study in detail about various techniques for disease identification and classification is also done.

There are an increasing number of papers published every year. It is desired for researchers to find the new high-quality papers, which is also a challenging task due to the lack of citation information. In this paper, we propose a novel method to predicate a new paper influence by collaboratively learning the latent vectors of paper features and correlations. We propose the concept topic related authority to integrate the dynamic topic model with paper citations so as to learn how content and authors influence a paper quality. We adopt the Factorization Machine method to collaboratively learn the latent vectors of correlations between different paper features. Comparing with traditional methods, it does not require the citation information to evaluate a paper quality, which is appropriate for new published papers. We conduct extensive evaluation against a real dataset crawled from ACM Digital Library. The results show that our method outperforms the other methods.

Recycled paper is extensively used worldwide. In the last decades its market has expanded considerably. The increasing use of recycled paper in papermaking has led to the production of paper containing several types of impurities. Consequently, wastepaper mills are forced to implement quality control schemes for evaluating the incoming wastepaper stock, thus guarantying the specifications of the final product. The main objective of this work is to present a fast and reliable system for identifying different paper types. Therefore, undesirable paper types can be refused, improving the performance of the paper machine and the final quality of the paper manufactured. For this purpose two fast techniques, i.e., Fourier transform mid-infrared (FTIR) and reflectance near-infrared (*IR) were applied to acquire the infrared spectra of the paper samples. *ext, four processing multivariate methods, i.e., principal component analysis (PCA), canonical variate analysis (CVA), extended canonical variate analysis (ECVA) and support vector machines (SVM) were employed in the feature extraction –or dimension reduction– stage. Afterwards, the k nearest neighbors algorithm (k**) was used in the classification phase. Experimental results show the usefulness of the proposed methodology and the potential of both FTIR and *IR spectroscopic methods. Using the FTIR spectrum in association with SVM and k** the system achieved maximum classification accuracy of 100%, whereas using the *IR spectrum in association with ECVA or SVM and k** the system achieved maximum classification accuracy of 96.4%

For the image restoration problem of shredded paper broken by shredder with the same marginal feature, a new method based on spatial feature classification is proposed in this paper. Because no obvious marginal feature for rule shredder paper can be used to split, the feature similarity classification processing can greatly reduce the solution scale and improve the algorithm efficiency. The horizontal projection position of shredded paper image can be used to distinguish the same line for the shredded paper with words. Then the marginal moment invariants distance are defined by the left and right part spatial correlation information. The original order of adjacent shredded paper in the same line can be ensured by the minimum marginal distance. Similarly ordering up and down can be completed by the minimum marginal distance among up-down part stitching images. For the image stitching and restoration of large-scale rule shredded paper, the actual experimental result shows the good stability, efficiency and strong robustness.

Hand gesture recognition system has a good attention now days because of easy interaction between human and machine. The focus of developing hand gesture is to create a better communication between human and computer for conveying information. This paper presents a survey on recent technology of hand gesture recognition both static and dynamic. It shows all method that was used for hand gesture recognition in different research paper.

For the restoration problem of shredded paper broken by shredder machines with the same marginal feature, a new method based on ant colony algorithm with classification is proposed in this paper. Firstly, shredded paper feature vector can be extracted by image space information. Secondly, the similar matrix and the marginal distance matrix are defined by the feature vector and left-right part image information respectively for every pieces of paper. At last, the original order of shredded paper can be ensured by ant colony algorithm depending on the minimum marginal distance sum among adjacent pieces of paper. For the stitching and restoration of large-scale broken pieces of paper without obvious marginal feature, the classification processing in this paper can greatly reduce the solution scale and improve the algorithm efficiency. The actual experimental result shows the good stability, efficiency and strong robustness of the algorithm.

As part of a module re-unification project of an industrial partner's code, spanning one systems and two derivative systems, the feature-clone variants across these systems have to be extracted, to be later re-unified as singular code elements for re-use. To assist developers with this task, the CoRA (The Code Re-unification Application) tool was designed and implemented. An approach, and the subsequent design of the tool was derived from reflection on manual feature-location/clonedetection efforts on the company's systems, in the first phase of an action research cycle where the approach/implementation will be iteratively trialled, and subsequently refined, in-situ. A pilot study is discussed that leads to the proposed tool. The tool combines a hybrid (textual-static) feature location technique and a textual clone detection technique for featureclone identification. In this paper, the rationale behind the CoRA tool is presented, followed by a tool overview and its implementation details. Finally, an example use case shows how the tool is used to locate clones of a particular feature.

A code clone is a pair of code fragments, within or between software systems that are similar. Since code clones often negatively impact the maintainability of a software system, a great many numbers of code clone detection techniques and tools have been proposed and studied over the last decade. To detect all possible similar source code patterns in general, the clone detection tools work on syntax level (such as texts, tokens, AST and so on) while lacking user-specific preferences. This often means the reported clones must be manually validated prior to any analysis in order to filter out the true positive clones from task or user-specific considerations. This manual clone validation effort is very time-consuming and often error-prone, in particular for large-scale clone detection. In this paper, we propose a machine learning based approach for automating the validation process. In an experiment with clones detected by several clone detectors in several different software systems, we found our approach has an accuracy of up to 87.4% when compared against the manual validation by multiple expert judges. The proposed method shows promising results in several comparative studies with the existing related approaches for automatic code clone validation. We also present our experimental results in terms of different code clone detection tools, machine learning algorithms and open source software systems.

The aim of this research is to classify the front side (observe) and back side (reserve) of Myanmar currencies (kyats) in three denominations. The system consists of binary image conversion, noise removal, feature extraction, and recognition. Zernike moment is used for feature extraction. The features from the top left, top right, bottom left and bottom right of the currency are extracted. The classification of the correct currency using the k -nearest neighbor algorithm based on the data set generated by Zernike Moment. The accuracy is measured based on the confusion matrix. The experimental results suggested that the approach is sufficiently effective and efficient in classifying the paper currencies.

The problem of classifying a research article into one or more fields of science is of particular importance for the academic search engines and digital libraries. A robust classification algorithm offers the users a wide variety of useful tools, such as the refinement of their search results, the browsing of articles by category, the recommendation of other similar articles, etc. In the current literature we encounter approaches which attempt to address this problem without taking into consideration important parameters such as the previous history of the authors and the categorization of the scientific journals which publish the articles. In addition, the existing works overlook the huge volume of the involved academic data. In this paper, we expand an existing effective algorithm for research articles classification, and we parallelize it on Apache Spark –a parallelization framework which is capable of sharing large amounts of data into the main memory of the nodes of a cluster– to enable the processing of large academic datasets. Furthermore, we present data manipulation methodologies which are useful not only for this particular problem, but also for most parallel machine learning approaches. In our experimental evaluation, we demonstrate that our proposed algorithm is considerably more accurate than the supervised learning approaches implemented within the machine learning library of Spark, whereas it outperforms them in terms of execution speed by a significant margin.

The characteristics of currency note vary from country to country. The recognition of currency depends on the characteristics of currency note of a particular country. Due to use for a long time, currency note may be contaminated by noises. It is difficult for a system to recognize old, torn, and noisy images of currency. This work focuses on Bangladeshi paper currency recognition based on supervised learning. Here a recognition system for Bangladeshi banknotes has been proposed. Initially, it receives the images of Bangladeshi banknotes as input. Then each image is divided into three channels. Filtering is applied to each channel. Finally, the red, the green, and the blue channels are recombined to get back the RGB image. Different features such as HSV, edge, and grey-level co-occurrence matrix are extracted from the RGB image. Total features are stored. Then Euclidean distance of features between the input image and the template images are computed. The minimum distance provides the required output. The performance of the proposed system has been evaluated using a challenging dataset with a wide variety of conditions, such as old, dirty, and torn banknotes.

This paper presents a method for identifying the most similar image in the visible image library and the given infrared image. First the method of visible and infrared image edge extraction respectively, and then use the perceptual hash encoding method of edge image, and get the image similarity by comparing the two image hash encoding Hamming distance. As the infrared image noise more perceptual hash algorithm based on pixel values may have errors of judgment, so the use of hash algorithm to get similarity on the 3 visible images using the SURF algorithm to compare the similarity. Based on the slope consistency, two image matching points are obtained. The most visible images of matching points are selected as the most similar images to infrared images. Experimental results show that the method can identify the visible image with the highest similarity between infrared image and visible image library.

In the last decade, data security has become a primary concern for an increasing amount of companies around the world. Protecting the customer's privacy is now at the core of many businesses operating in any kind of market. Thus, the demand for new technologies to safeguard user data and prevent data breaches has increased accordingly. In this work, we investigate a machine learning-based approach to automatically extract sources and sinks from arbitrary Java libraries. Our method exploits several different features based on semantic, syntactic, intra-procedural dataflow and class-hierarchy traits embedded into the bytecode to distinguish sources and sinks. The performed experiments show that, under certain conditions and after some preprocessing, sources and sinks across different libraries share common characteristics that allow a machine learning model to distinguish them from the other library methods. The prototype model achieved remarkable results of 86% accuracy and 81% F-measure on our validation set of roughly 600 methods.

As the presentation slides are of vital importance in a person's career, a significant amount of time is spent for its preparation. An automatic paper-summarizer will reduce the amount of time and human effort. Presently, tools exist for formatting and designing themes for slides, but not for content generation. This paper proposes a summarization system that automatically generates presentation slides from scientific papers. The proposed system accepts scientific papers as input and prepares presentation slides, which includes text, images and graphs present in them. Papers and slides are learned and trained using Support Vector Regression model. It calculates and predicts the regression scores. Based on the regression scores, the relevant content of the paper is selected for generating summary. The generated summary is used for making slides using Integer Linear Programming along with the images present in papers.

Text extraction is a crucial stage of analyzing Journal papers. Journal papers generally are in PDF format which is semi structured data. Journal papers are presented into different sections like Introduction, Methodology, Experimental setup, Result and analysis etc. so that it is easy to access information from any section as per the reader's interest. The main importance on section extraction is to find a representative subset of the data, which contains the information of the entire set. Various approaches to extract sections from research papers include stastical methods, NLP, Machine Learning etc. In this paper we present review of various extraction techniques from a PDF document.

Based on the analysis of the types and characteristics of Chinese traditional paper-cut patterns, we propose a new algorithm for extracting and vectoring the patterns. On the contour extraction, the algorithm uses a grayscale, binary, contour tracking method. On the vectorization, we improved the segment Displaced curve segments so that you can make the final contour paper cutting pattern vector into a control point and a curve. Using this algorithm, we can quickly get the vector contour of the pattern of manual cutting. Further, we can carry on the processing, such as filling, deformation, combination, animation control, so as to complete the core function of the paper cutting animation software. The algorithm is also suitable for the extraction and the vector of common bitmap image, although the algorithm is put forward for the particularity of the paper cutting patterns.

Platforms for publishing research papers are increasing largely that contribute to big data as their volume is humongous and are unstandardized. Classification of this huge chunk of data is one of the biggest challenges in Information Retrieval. In this paper we discuss a scoring based unsupervised learning approach to extract relevant features and classify the research papers according to their content on a two class dataset. Feature extraction is carried out by analyzing the sections of a research paper and scoring them, followed by a two level hybrid classification technique based on title and conceptual summary using graph clustering. Promising experimental results are observed for data set using our research paper classification method. We present the experimental results of our proposed algorithm for feature extraction and clustering and compare the same with different approaches reported in literature.

Advancement of technology over the recent past has led to an increase in circulation of counterfeit notes in today's economy. To combat this issue, it is essential that there exists an efficient mechanism to detect fake banknotes. The main problem with current systems is the trade-off between speed and complexity. This paper proposes a system that can classify and subsequently verify Indian paper currency using fundamental image processing techniques. It uses the comparison between the input banknote and the calculated reference values for different parameters of original banknotes in a similar environment. This system maintains its simplicity while still having high accuracy of 100% for classification and 90% for validity verification.

In this paper, we investigate the geometrical shape of cut-out patterns and the classification techniques, then introduce the six geometry features definition, including shape-factor, complexity, extendability, eccentricity, solidity and modal-ratio and propose the application of BP neural networks to train, classify and identify the patterns. The proposed scheme has the advantages of classifying and identifying the excessive geometrical artistic deformations. Experimental results demonstrate the superiority of the pattern recognition and the algorithm is simpler and easier to implement.

Exploiting resource reusability and low precision in neural networks is a promising approach to achieve energy efficient computational platforms. This research presents two generalizable approaches to reuse resources in feed-forward neural networks and demonstrated on extreme learning machines. In the first approach, coalescing, a single stack of neuronal units perform both feature extraction and classification tasks through shared resources. In the second approach, folding, the neurons in a high-dimension feedforward layer are folded to execute multiple-tasks. The folding technique can also be combined with low precision modules. The proposed design techniques are validated for a classification task on binary (Australian credit and Diabetes corpus) and multi-class (MNIST) dataset. The total power consumption is measured to be 3.65 mW on TSMC 65nm technology node, while yielding an accuracy of 91.7% for MNIST.

Research papers are often referred by researchers. But finding the desired or relevant research paper quickly and accurately is very difficult. As there are lot of research papers in given dataset and keeps enormously increasing. There is a need to avail automated processing approach for tackling such a huge volume of dataset to retrieve relevant papers accurately. This paper proposes a framework which works in two phases. In first phase, papers are stored in a dataset. Initially the Information Gain is calculated for finding features of each document in dataset. These features of document are used for classification. For classification of document, Naive Bayes approach is used. In second phase, relevant documents are retrieved based on search query. In this features of search query are extracted using Information Gain. The extracted features are used for retrieving relevant research papers. The extracted features of query and features of document stored in a dataset are compared using cosine similarity approach.

In this paper, a framework for human hand movement recognition is implemented for humans to play rock-paper-scissors game interactively against computer using a webcam. The game is the classic rock-paper-scissors game; however, first player is human and second player is computer. Human hand movements are recognized using a webcam, and the game is fully controlled using hand gestures. A dataset is constructed in this work by collecting hand images of different people. The dataset contains rock, paper, and scissors movements of different humanbeings. There are 8 photos for each movement and person using their left and right hands. The dataset is constructed using a controlled environment with black background and close up hand images. To recognize player movements two different feature extraction methods and two different classification methods are used and compared. Histogram of oriented gradients feature extraction method combined with support vector machines are found to be effective. It presents fast and effective computer vision framework for seamless playing experience with a sensitivity of %92.9.

Falsification of medicines is a big problem in many developing countries, where technological infrastructure is inadequate to detect these harmful products. We have developed a set of inexpensive paper cards, called Paper Analytical Devices (PADs), which can efficiently classify drugs based on their chemical composition, as a potential solution to the problem. These cards have different reagents embedded in them which produce a set of distinctive color descriptors upon reacting with the chemical compounds that constitute pharmaceutical dosage forms. If a falsified version of the medicine lacks the active ingredient or includes substitute fillers, the difference in color is perceivable by humans. However, reading the cards with accuracy takes training and practice, which may hamper their scaling and implementation in low resource settings. To deal with this, we have developed an automatic visual recognition system to read the results from the PAD images. At first, the optimal set of reagents was found by running singular value decomposition on the intensity values of the color tones in the card images. A dataset of cards embedded with these reagents is produced to generate the most distinctive results for a set of 26 different active pharmaceutical ingredients (APIs) and excipients. Then, we train two popular convolutional neural network (CNN) models, with the card images. We also extract some "hand-crafted" features from the images and train a nearest neighbor classifier and a non-linear support vector machine with them. On testing, higher-level features performed much better in accurately classifying the PAD images, with the CNN models reaching the highest average accuracy of over 94\%., Comment: in Proc. IEEE Winter Conference on Applications of Computer Vision (WACV), 2016