Search results for: scene text recognition
Commenced in January 2007
Frequency: Monthly
Edition: International
Paper Count: 3004

Search results for: scene text recognition

2734 Makhraj Recognition Using Convolutional Neural Network

Authors: Zan Azma Nasruddin, Irwan Mazlin, Nor Aziah Daud, Fauziah Redzuan, Fariza Hanis Abdul Razak

Abstract:

This paper focuses on a machine learning that learn the correct pronunciation of Makhraj Huroofs. Usually, people need to find an expert to pronounce the Huroof accurately. In this study, the researchers have developed a system that is able to learn the selected Huroofs which are ha, tsa, zho, and dza using the Convolutional Neural Network. The researchers present the chosen type of the CNN architecture to make the system that is able to learn the data (Huroofs) as quick as possible and produces high accuracy during the prediction. The researchers have experimented the system to measure the accuracy and the cross entropy in the training process.

Keywords: convolutional neural network, Makhraj recognition, speech recognition, signal processing, tensorflow

Procedia PDF Downloads 289
2733 Multilabel Classification with Neural Network Ensemble Method

Authors: Sezin Ekşioğlu

Abstract:

Multilabel classification has a huge importance for several applications, it is also a challenging research topic. It is a kind of supervised learning that contains binary targets. The distance between multilabel and binary classification is having more than one class in multilabel classification problems. Features can belong to one class or many classes. There exists a wide range of applications for multi label prediction such as image labeling, text categorization, gene functionality. Even though features are classified in many classes, they may not always be properly classified. There are many ensemble methods for the classification. However, most of the researchers have been concerned about better multilabel methods. Especially little ones focus on both efficiency of classifiers and pairwise relationships at the same time in order to implement better multilabel classification. In this paper, we worked on modified ensemble methods by getting benefit from k-Nearest Neighbors and neural network structure to address issues within a beneficial way and to get better impacts from the multilabel classification. Publicly available datasets (yeast, emotion, scene and birds) are performed to demonstrate the developed algorithm efficiency and the technique is measured by accuracy, F1 score and hamming loss metrics. Our algorithm boosts benchmarks for each datasets with different metrics.

Keywords: multilabel, classification, neural network, KNN

Procedia PDF Downloads 115
2732 A Semantic and Concise Structure to Represent Human Actions

Authors: Tobias Strübing, Fatemeh Ziaeetabar

Abstract:

Humans usually manipulate objects with their hands. To represent these actions in a simple and understandable way, we need to use a semantic framework. For this purpose, the Semantic Event Chain (SEC) method has already been presented which is done by consideration of touching and non-touching relations between manipulated objects in a scene. This method was improved by a computational model, the so-called enriched Semantic Event Chain (eSEC), which incorporates the information of static (e.g. top, bottom) and dynamic spatial relations (e.g. moving apart, getting closer) between objects in an action scene. This leads to a better action prediction as well as the ability to distinguish between more actions. Each eSEC manipulation descriptor is a huge matrix with thirty rows and a massive set of the spatial relations between each pair of manipulated objects. The current eSEC framework has so far only been used in the category of manipulation actions, which eventually involve two hands. Here, we would like to extend this approach to a whole body action descriptor and make a conjoint activity representation structure. For this purpose, we need to do a statistical analysis to modify the current eSEC by summarizing while preserving its features, and introduce a new version called Enhanced eSEC or (e2SEC). This summarization can be done from two points of the view: 1) reducing the number of rows in an eSEC matrix, 2) shrinking the set of possible semantic spatial relations. To achieve these, we computed the importance of each matrix row in an statistical way, to see if it is possible to remove a particular one while all manipulations are still distinguishable from each other. On the other hand, we examined which semantic spatial relations can be merged without compromising the unity of the predefined manipulation actions. Therefore by performing the above analyses, we made the new e2SEC framework which has 20% fewer rows, 16.7% less static spatial and 11.1% less dynamic spatial relations. This simplification, while preserving the salient features of a semantic structure in representing actions, has a tremendous impact on the recognition and prediction of complex actions, as well as the interactions between humans and robots. It also creates a comprehensive platform to integrate with the body limbs descriptors and dramatically increases system performance, especially in complex real time applications such as human-robot interaction prediction.

Keywords: enriched semantic event chain, semantic action representation, spatial relations, statistical analysis

Procedia PDF Downloads 72
2731 Grammatical and Lexical Cohesion in the Japan’s Prime Minister Shinzo Abe’s Speech Text ‘Nihon wa Modottekimashita’

Authors: Nadya Inda Syartanti

Abstract:

This research aims to identify, classify, and analyze descriptively the aspects of grammatical and lexical cohesion in the speech text of Japan’s Prime Minister Shinzo Abe entitled Nihon wa Modotte kimashita delivered in Washington DC, the United States on February 23, 2013, as a research data source. The method used is qualitative research, which uses descriptions through words that are applied by analyzing aspects of grammatical and lexical cohesion proposed by Halliday and Hasan (1976). The aspects of grammatical cohesion consist of references (personal, demonstrative, interrogative pronouns), substitution, ellipsis, and conjunction. In contrast, lexical cohesion consists of reiteration (repetition, synonym, antonym, hyponym, meronym) and collocation. Data classification is based on the 6 aspects of the cohesion. Through some aspects of cohesion, this research tries to find out the frequency of using grammatical and lexical cohesion in Shinzo Abe's speech text entitled Nihon wa Modotte kimashita. The results of this research are expected to help overcome the difficulty of understanding speech texts in Japanese. Therefore, this research can be a reference for learners, researchers, and anyone who is interested in the field of discourse analysis.

Keywords: cohesion, grammatical cohesion, lexical cohesion, speech text, Shinzo Abe

Procedia PDF Downloads 120
2730 Spontaneous Message Detection of Annoying Situation in Community Networks Using Mining Algorithm

Authors: P. Senthil Kumari

Abstract:

Main concerns in data mining investigation are social controls of data mining for handling ambiguity, noise, or incompleteness on text data. We describe an innovative approach for unplanned text data detection of community networks achieved by classification mechanism. In a tangible domain claim with humble secrecy backgrounds provided by community network for evading annoying content is presented on consumer message partition. To avoid this, mining methodology provides the capability to unswervingly switch the messages and similarly recover the superiority of ordering. Here we designated learning-centered mining approaches with pre-processing technique to complete this effort. Our involvement of work compact with rule-based personalization for automatic text categorization which was appropriate in many dissimilar frameworks and offers tolerance value for permits the background of comments conferring to a variety of conditions associated with the policy or rule arrangements processed by learning algorithm. Remarkably, we find that the choice of classifier has predicted the class labels for control of the inadequate documents on community network with great value of effect.

Keywords: text mining, data classification, community network, learning algorithm

Procedia PDF Downloads 467
2729 The Artificial Intelligence Technologies Used in PhotoMath Application

Authors: Tala Toonsi, Marah Alagha, Lina Alnowaiser, Hala Rajab

Abstract:

This report is about the Photomath app, which is an AI application that uses image recognition technology, specifically optical character recognition (OCR) algorithms. The (OCR) algorithm translates the images into a mathematical equation, and the app automatically provides a step-by-step solution. The application supports decimals, basic arithmetic, fractions, linear equations, and multiple functions such as logarithms. Testing was conducted to examine the usage of this app, and results were collected by surveying ten participants. Later, the results were analyzed. This paper seeks to answer the question: To what level the artificial intelligence features are accurate and the speed of process in this app. It is hoped this study will inform about the efficiency of AI in Photomath to the users.

Keywords: photomath, image recognition, app, OCR, artificial intelligence, mathematical equations.

Procedia PDF Downloads 126
2728 A Human Activity Recognition System Based on Sensory Data Related to Object Usage

Authors: M. Abdullah, Al-Wadud

Abstract:

Sensor-based activity recognition systems usually accounts which sensors have been activated to perform an activity. The system then combines the conditional probabilities of those sensors to represent different activities and takes the decision based on that. However, the information about the sensors which are not activated may also be of great help in deciding which activity has been performed. This paper proposes an approach where the sensory data related to both usage and non-usage of objects are utilized to make the classification of activities. Experimental results also show the promising performance of the proposed method.

Keywords: Naïve Bayesian, based classification, activity recognition, sensor data, object-usage model

Procedia PDF Downloads 279
2727 Wasting Human and Computer Resources

Authors: Mária Csernoch, Piroska Biró

Abstract:

The legends about “user-friendly” and “easy-to-use” birotical tools (computer-related office tools) have been spreading and misleading end-users. This approach has led us to the extremely high number of incorrect documents, causing serious financial losses in the creating, modifying, and retrieving processes. Our research proved that there are at least two sources of this underachievement: (1) The lack of the definition of the correctly edited, formatted documents. Consequently, end-users do not know whether their methods and results are correct or not. They are not aware of their ignorance. They are so ignorant that their ignorance does not allow them to realize their lack of knowledge. (2) The end-users’ problem-solving methods. We have found that in non-traditional programming environments end-users apply, almost exclusively, surface approach metacognitive methods to carry out their computer related activities, which are proved less effective than deep approach methods. Based on these findings we have developed deep approach methods which are based on and adapted from traditional programming languages. In this study, we focus on the most popular type of birotical documents, the text-based documents. We have provided the definition of the correctly edited text, and based on this definition, adapted the debugging method known in programming. According to the method, before the realization of text editing, a thorough debugging of already existing texts and the categorization of errors are carried out. With this method in advance to real text editing users learn the requirements of text-based documents and also of the correctly formatted text. The method has been proved much more effective than the previously applied surface approach methods. The advantages of the method are that the real text handling requires much less human and computer sources than clicking aimlessly in the GUI (Graphical User Interface), and the data retrieval is much more effective than from error-prone documents.

Keywords: deep approach metacognitive methods, error-prone birotical documents, financial losses, human and computer resources

Procedia PDF Downloads 350
2726 Performance Assessment of Multi-Level Ensemble for Multi-Class Problems

Authors: Rodolfo Lorbieski, Silvia Modesto Nassar

Abstract:

Many supervised machine learning tasks require decision making across numerous different classes. Multi-class classification has several applications, such as face recognition, text recognition and medical diagnostics. The objective of this article is to analyze an adapted method of Stacking in multi-class problems, which combines ensembles within the ensemble itself. For this purpose, a training similar to Stacking was used, but with three levels, where the final decision-maker (level 2) performs its training by combining outputs from the tree-based pair of meta-classifiers (level 1) from Bayesian families. These are in turn trained by pairs of base classifiers (level 0) of the same family. This strategy seeks to promote diversity among the ensembles forming the meta-classifier level 2. Three performance measures were used: (1) accuracy, (2) area under the ROC curve, and (3) time for three factors: (a) datasets, (b) experiments and (c) levels. To compare the factors, ANOVA three-way test was executed for each performance measure, considering 5 datasets by 25 experiments by 3 levels. A triple interaction between factors was observed only in time. The accuracy and area under the ROC curve presented similar results, showing a double interaction between level and experiment, as well as for the dataset factor. It was concluded that level 2 had an average performance above the other levels and that the proposed method is especially efficient for multi-class problems when compared to binary problems.

Keywords: stacking, multi-layers, ensemble, multi-class

Procedia PDF Downloads 229
2725 Short Text Classification for Saudi Tweets

Authors: Asma A. Alsufyani, Maram A. Alharthi, Maha J. Althobaiti, Manal S. Alharthi, Huda Rizq

Abstract:

Twitter is one of the most popular microblogging sites that allows users to publish short text messages called 'tweets'. Increasing the number of accounts to follow (followings) increases the number of tweets that will be displayed from different topics in an unclassified manner in the timeline of the user. Therefore, it can be a vital solution for many Twitter users to have their tweets in a timeline classified into general categories to save the user’s time and to provide easy and quick access to tweets based on topics. In this paper, we developed a classifier for timeline tweets trained on a dataset consisting of 3600 tweets in total, which were collected from Saudi Twitter and annotated manually. We experimented with the well-known Bag-of-Words approach to text classification, and we used support vector machines (SVM) in the training process. The trained classifier performed well on a test dataset, with an average F1-measure equal to 92.3%. The classifier has been integrated into an application, which practically proved the classifier’s ability to classify timeline tweets of the user.

Keywords: corpus creation, feature extraction, machine learning, short text classification, social media, support vector machine, Twitter

Procedia PDF Downloads 112
2724 Moral Wrongdoers: Evaluating the Value of Moral Actions Performed by War Criminals

Authors: Jean-Francois Caron

Abstract:

This text explores the value of moral acts performed by war criminals, and the extent to which they should alleviate the punishment these individuals ought to receive for violating the rules of war. Without neglecting the necessity of retribution in war crimes cases, it argues from an ethical perspective that we should not rule out the possibility of considering lesser punishments for war criminals who decide to perform a moral act, as it might produce significant positive moral outcomes. This text also analyzes how such a norm could be justified from a moral perspective.

Keywords: war criminals, pardon, amnesty, retribution

Procedia PDF Downloads 241
2723 A Deep Learning Approach to Subsection Identification in Electronic Health Records

Authors: Nitin Shravan, Sudarsun Santhiappan, B. Sivaselvan

Abstract:

Subsection identification, in the context of Electronic Health Records (EHRs), is identifying the important sections for down-stream tasks like auto-coding. In this work, we classify the text present in EHRs according to their information, using machine learning and deep learning techniques. We initially describe briefly about the problem and formulate it as a text classification problem. Then, we discuss upon the methods from the literature. We try two approaches - traditional feature extraction based machine learning methods and deep learning methods. Through experiments on a private dataset, we establish that the deep learning methods perform better than the feature extraction based Machine Learning Models.

Keywords: deep learning, machine learning, semantic clinical classification, subsection identification, text classification

Procedia PDF Downloads 172
2722 Features Vector Selection for the Recognition of the Fragmented Handwritten Numeric Chains

Authors: Salim Ouchtati, Aissa Belmeguenai, Mouldi Bedda

Abstract:

In this study, we propose an offline system for the recognition of the fragmented handwritten numeric chains. Firstly, we realized a recognition system of the isolated handwritten digits, in this part; the study is based mainly on the evaluation of neural network performances, trained with the gradient backpropagation algorithm. The used parameters to form the input vector of the neural network are extracted from the binary images of the isolated handwritten digit by several methods: the distribution sequence, sondes application, the Barr features, and the centered moments of the different projections and profiles. Secondly, the study is extended for the reading of the fragmented handwritten numeric chains constituted of a variable number of digits. The vertical projection was used to segment the numeric chain at isolated digits and every digit (or segment) was presented separately to the entry of the system achieved in the first part (recognition system of the isolated handwritten digits).

Keywords: features extraction, handwritten numeric chains, image processing, neural networks

Procedia PDF Downloads 229
2721 Identification of Text Domains and Register Variation through the Analysis of Lexical Distribution in a Bangla Mass Media Text Corpus

Authors: Mahul Bhattacharyya, Niladri Sekhar Dash

Abstract:

The present research paper is an experimental attempt to investigate the nature of variation in the register in three major text domains, namely, social, cultural, and political texts collected from the corpus of Bangla printed mass media texts. This present study uses a corpus of a moderate amount of Bangla mass media text that contains nearly one million words collected from different media sources like newspapers, magazines, advertisements, periodicals, etc. The analysis of corpus data reveals that each text has certain lexical properties that not only control their identity but also mark their uniqueness across the domains. At first, the subject domains of the texts are classified into two parameters namely, ‘Genre' and 'Text Type'. Next, some empirical investigations are made to understand how the domains vary from each other in terms of lexical properties like both function and content words. Here the method of comparative-cum-contrastive matching of lexical load across domains is invoked through word frequency count to track how domain-specific words and terms may be marked as decisive indicators in the act of specifying the textual contexts and subject domains. The study shows that the common lexical stock that percolates across all text domains are quite dicey in nature as their lexicological identity does not have any bearing in the act of specifying subject domains. Therefore, it becomes necessary for language users to anchor upon certain domain-specific lexical items to recognize a text that belongs to a specific text domain. The eventual findings of this study confirm that texts belonging to different subject domains in Bangla news text corpus clearly differ on the parameters of lexical load, lexical choice, lexical clustering, lexical collocation. In fact, based on these parameters, along with some statistical calculations, it is possible to classify mass media texts into different types to mark their relation with regard to the domains they should actually belong. The advantage of this analysis lies in the proper identification of the linguistic factors which will give language users a better insight into the method they employ in text comprehension, as well as construct a systemic frame for designing text identification strategy for language learners. The availability of huge amount of Bangla media text data is useful for achieving accurate conclusions with a certain amount of reliability and authenticity. This kind of corpus-based analysis is quite relevant for a resource-poor language like Bangla, as no attempt has ever been made to understand how the structure and texture of Bangla mass media texts vary due to certain linguistic and extra-linguistic constraints that are actively operational to specific text domains. Since mass media language is assumed to be the most 'recent representation' of the actual use of the language, this study is expected to show how the Bangla news texts reflect the thoughts of the society and how they leave a strong impact on the thought process of the speech community.

Keywords: Bangla, corpus, discourse, domains, lexical choice, mass media, register, variation

Procedia PDF Downloads 141
2720 Semantic Data Schema Recognition

Authors: Aïcha Ben Salem, Faouzi Boufares, Sebastiao Correia

Abstract:

The subject covered in this paper aims at assisting the user in its quality approach. The goal is to better extract, mix, interpret and reuse data. It deals with the semantic schema recognition of a data source. This enables the extraction of data semantics from all the available information, inculding the data and the metadata. Firstly, it consists of categorizing the data by assigning it to a category and possibly a sub-category, and secondly, of establishing relations between columns and possibly discovering the semantics of the manipulated data source. These links detected between columns offer a better understanding of the source and the alternatives for correcting data. This approach allows automatic detection of a large number of syntactic and semantic anomalies.

Keywords: schema recognition, semantic data profiling, meta-categorisation, semantic dependencies inter columns

Procedia PDF Downloads 384
2719 Resource Framework Descriptors for Interestingness in Data

Authors: C. B. Abhilash, Kavi Mahesh

Abstract:

Human beings are the most advanced species on earth; it's all because of the ability to communicate and share information via human language. In today's world, a huge amount of data is available on the web in text format. This has also resulted in the generation of big data in structured and unstructured formats. In general, the data is in the textual form, which is highly unstructured. To get insights and actionable content from this data, we need to incorporate the concepts of text mining and natural language processing. In our study, we mainly focus on Interesting data through which interesting facts are generated for the knowledge base. The approach is to derive the analytics from the text via the application of natural language processing. Using semantic web Resource framework descriptors (RDF), we generate the triple from the given data and derive the interesting patterns. The methodology also illustrates data integration using the RDF for reliable, interesting patterns.

Keywords: RDF, interestingness, knowledge base, semantic data

Procedia PDF Downloads 116
2718 Speaker Identification by Atomic Decomposition of Learned Features Using Computational Auditory Scene Analysis Principals in Noisy Environments

Authors: Thomas Bryan, Veton Kepuska, Ivica Kostanic

Abstract:

Speaker recognition is performed in high Additive White Gaussian Noise (AWGN) environments using principals of Computational Auditory Scene Analysis (CASA). CASA methods often classify sounds from images in the time-frequency (T-F) plane using spectrograms or cochleargrams as the image. In this paper atomic decomposition implemented by matching pursuit performs a transform from time series speech signals to the T-F plane. The atomic decomposition creates a sparsely populated T-F vector in “weight space” where each populated T-F position contains an amplitude weight. The weight space vector along with the atomic dictionary represents a denoised, compressed version of the original signal. The arraignment or of the atomic indices in the T-F vector are used for classification. Unsupervised feature learning implemented by a sparse autoencoder learns a single dictionary of basis features from a collection of envelope samples from all speakers. The approach is demonstrated using pairs of speakers from the TIMIT data set. Pairs of speakers are selected randomly from a single district. Each speak has 10 sentences. Two are used for training and 8 for testing. Atomic index probabilities are created for each training sentence and also for each test sentence. Classification is performed by finding the lowest Euclidean distance between then probabilities from the training sentences and the test sentences. Training is done at a 30dB Signal-to-Noise Ratio (SNR). Testing is performed at SNR’s of 0 dB, 5 dB, 10 dB and 30dB. The algorithm has a baseline classification accuracy of ~93% averaged over 10 pairs of speakers from the TIMIT data set. The baseline accuracy is attributable to short sequences of training and test data as well as the overall simplicity of the classification algorithm. The accuracy is not affected by AWGN and produces ~93% accuracy at 0dB SNR.

Keywords: time-frequency plane, atomic decomposition, envelope sampling, Gabor atoms, matching pursuit, sparse dictionary learning, sparse autoencoder

Procedia PDF Downloads 250
2717 Speech Recognition Performance by Adults: A Proposal for a Battery for Marathi

Authors: S. B. Rathna Kumar, Pranjali A Ujwane, Panchanan Mohanty

Abstract:

The present study aimed to develop a battery for assessing speech recognition performance by adults in Marathi. A total of four word lists were developed by considering word frequency, word familiarity, words in common use, and phonemic balance. Each word list consists of 25 words (15 monosyllabic words in CVC structure and 10 monosyllabic words in CVCV structure). Equivalence analysis and performance-intensity function testing was carried using the four word lists on a total of 150 native speakers of Marathi belonging to different regions of Maharashtra (Vidarbha, Marathwada, Khandesh and Northern Maharashtra, Pune, and Konkan). The subjects were further equally divided into five groups based on above mentioned regions. It was found that there was no significant difference (p > 0.05) in the speech recognition performance between groups for each word list and between word lists for each group. Hence, the four word lists developed were equally difficult for all the groups and can be used interchangeably. The performance-intensity (PI) function curve showed semi-linear function, and the groups’ mean slope of the linear portions of the curve indicated an average linear slope of 4.64%, 4.73%, 4.68%, and 4.85% increase in word recognition score per dB for list 1, list 2, list 3 and list 4 respectively. Although, there is no data available on speech recognition tests for adults in Marathi, most of the findings of the study are in line with the findings of research reports on other languages. The four word lists, thus developed, were found to have sufficient reliability and validity in assessing speech recognition performance by adults in Marathi.

Keywords: speech recognition performance, phonemic balance, equivalence analysis, performance-intensity function testing, reliability, validity

Procedia PDF Downloads 320
2716 Face Recognition Using Body-Worn Camera: Dataset and Baseline Algorithms

Authors: Ali Almadan, Anoop Krishnan, Ajita Rattani

Abstract:

Facial recognition is a widely adopted technology in surveillance, border control, healthcare, banking services, and lately, in mobile user authentication with Apple introducing “Face ID” moniker with iPhone X. A lot of research has been conducted in the area of face recognition on datasets captured by surveillance cameras, DSLR, and mobile devices. Recently, face recognition technology has also been deployed on body-worn cameras to keep officers safe, enabling situational awareness and providing evidence for trial. However, limited academic research has been conducted on this topic so far, without the availability of any publicly available datasets with a sufficient sample size. This paper aims to advance research in the area of face recognition using body-worn cameras. To this aim, the contribution of this work is two-fold: (1) collection of a dataset consisting of a total of 136,939 facial images of 102 subjects captured using body-worn cameras in in-door and daylight conditions and (2) evaluation of various deep-learning architectures for face identification on the collected dataset. Experimental results suggest a maximum True Positive Rate(TPR) of 99.86% at False Positive Rate(FPR) of 0.000 obtained by SphereFace based deep learning architecture in daylight condition. The collected dataset and the baseline algorithms will promote further research and development. A downloadable link of the dataset and the algorithms is available by contacting the authors.

Keywords: face recognition, body-worn cameras, deep learning, person identification

Procedia PDF Downloads 125
2715 Pre-Analysis of Printed Circuit Boards Based on Multispectral Imaging for Vision Based Recognition of Electronics Waste

Authors: Florian Kleber, Martin Kampel

Abstract:

The increasing demand of gallium, indium and rare-earth elements for the production of electronics, e.g. solid state-lighting, photovoltaics, integrated circuits, and liquid crystal displays, will exceed the world-wide supply according to current forecasts. Recycling systems to reclaim these materials are not yet in place, which challenges the sustainability of these technologies. This paper proposes a multispectral imaging system as a basis for a vision based recognition system for valuable components of electronics waste. Multispectral images intend to enhance the contrast of images of printed circuit boards (single components, as well as labels) for further analysis, such as optical character recognition and entire printed circuit board recognition. The results show that a higher contrast is achieved in the near infrared compared to ultraviolet and visible light.

Keywords: electronics waste, multispectral imaging, printed circuit boards, rare-earth elements

Procedia PDF Downloads 380
2714 The Combination of the Mel Frequency Cepstral Coefficients, Perceptual Linear Prediction, Jitter and Shimmer Coefficients for the Improvement of Automatic Recognition System for Dysarthric Speech

Authors: Brahim Fares Zaidi

Abstract:

Our work aims to improve our Automatic Recognition System for Dysarthria Speech based on the Hidden Models of Markov and the Hidden Markov Model Toolkit to help people who are sick. With pronunciation problems, we applied two techniques of speech parameterization based on Mel Frequency Cepstral Coefficients and Perceptual Linear Prediction and concatenated them with JITTER and SHIMMER coefficients in order to increase the recognition rate of a dysarthria speech. For our tests, we used the NEMOURS database that represents speakers with dysarthria and normal speakers.

Keywords: ARSDS, HTK, HMM, MFCC, PLP

Procedia PDF Downloads 64
2713 A Framework of Product Information Service System Using Mobile Image Retrieval and Text Mining Techniques

Authors: Mei-Yi Wu, Shang-Ming Huang

Abstract:

The online shoppers nowadays often search the product information on the Internet using some keywords of products. To use this kind of information searching model, shoppers should have a preliminary understanding about their interesting products and choose the correct keywords. However, if the products are first contact (for example, the worn clothes or backpack of passengers which you do not have any idea about the brands), these products cannot be retrieved due to insufficient information. In this paper, we discuss and study the applications in E-commerce using image retrieval and text mining techniques. We design a reasonable E-commerce application system containing three layers in the architecture to provide users product information. The system can automatically search and retrieval similar images and corresponding web pages on Internet according to the target pictures which taken by users. Then text mining techniques are applied to extract important keywords from these retrieval web pages and search the prices on different online shopping stores with these keywords using a web crawler. Finally, the users can obtain the product information including photos and prices of their favorite products. The experiments shows the efficiency of proposed system.

Keywords: mobile image retrieval, text mining, product information service system, online marketing

Procedia PDF Downloads 319
2712 Multimodal Data Fusion Techniques in Audiovisual Speech Recognition

Authors: Hadeer M. Sayed, Hesham E. El Deeb, Shereen A. Taie

Abstract:

In the big data era, we are facing a diversity of datasets from different sources in different domains that describe a single life event. These datasets consist of multiple modalities, each of which has a different representation, distribution, scale, and density. Multimodal fusion is the concept of integrating information from multiple modalities in a joint representation with the goal of predicting an outcome through a classification task or regression task. In this paper, multimodal fusion techniques are classified into two main classes: model-agnostic techniques and model-based approaches. It provides a comprehensive study of recent research in each class and outlines the benefits and limitations of each of them. Furthermore, the audiovisual speech recognition task is expressed as a case study of multimodal data fusion approaches, and the open issues through the limitations of the current studies are presented. This paper can be considered a powerful guide for interested researchers in the field of multimodal data fusion and audiovisual speech recognition particularly.

Keywords: multimodal data, data fusion, audio-visual speech recognition, neural networks

Procedia PDF Downloads 70
2711 Distant Speech Recognition Using Laser Doppler Vibrometer

Authors: Yunbin Deng

Abstract:

Most existing applications of automatic speech recognition relies on cooperative subjects at a short distance to a microphone. Standoff speech recognition using microphone arrays can extend the subject to sensor distance somewhat, but it is still limited to only a few feet. As such, most deployed applications of standoff speech recognitions are limited to indoor use at short range. Moreover, these applications require air passway between the subject and the sensor to achieve reasonable signal to noise ratio. This study reports long range (50 feet) automatic speech recognition experiments using a Laser Doppler Vibrometer (LDV) sensor. This study shows that the LDV sensor modality can extend the speech acquisition standoff distance far beyond microphone arrays to hundreds of feet. In addition, LDV enables 'listening' through the windows for uncooperative subjects. This enables new capabilities in automatic audio and speech intelligence, surveillance, and reconnaissance (ISR) for law enforcement, homeland security and counter terrorism applications. The Polytec LDV model OFV-505 is used in this study. To investigate the impact of different vibrating materials, five parallel LDV speech corpora, each consisting of 630 speakers, are collected from the vibrations of a glass window, a metal plate, a plastic box, a wood slate, and a concrete wall. These are the common materials the application could encounter in a daily life. These data were compared with the microphone counterpart to manifest the impact of various materials on the spectrum of the LDV speech signal. State of the art deep neural network modeling approaches is used to conduct continuous speaker independent speech recognition on these LDV speech datasets. Preliminary phoneme recognition results using time-delay neural network, bi-directional long short term memory, and model fusion shows great promise of using LDV for long range speech recognition. To author’s best knowledge, this is the first time an LDV is reported for long distance speech recognition application.

Keywords: covert speech acquisition, distant speech recognition, DSR, laser Doppler vibrometer, LDV, speech intelligence surveillance and reconnaissance, ISR

Procedia PDF Downloads 141
2710 Interactive Shadow Play Animation System

Authors: Bo Wan, Xiu Wen, Lingling An, Xiaoling Ding

Abstract:

The paper describes a Chinese shadow play animation system based on Kinect. Users, without any professional training, can personally manipulate the shadow characters to finish a shadow play performance by their body actions and get a shadow play video through giving the record command to our system if they want. In our system, Kinect is responsible for capturing human movement and voice commands data. Gesture recognition module is used to control the change of the shadow play scenes. After packaging the data from Kinect and the recognition result from gesture recognition module, VRPN transmits them to the server-side. At last, the server-side uses the information to control the motion of shadow characters and video recording. This system not only achieves human-computer interaction, but also realizes the interaction between people. It brings an entertaining experience to users and easy to operate for all ages. Even more important is that the application background of Chinese shadow play embodies the protection of the art of shadow play animation.

Keywords: hadow play animation, Kinect, gesture recognition, VRPN, HCI

Procedia PDF Downloads 362
2709 The Effects of Three Pre-Reading Activities (Text Summary, Vocabulary Definition, and Pre-Passage Questions) on the Reading Comprehension of Iranian EFL Learners

Authors: Leila Anjomshoa, Firooz Sadighi

Abstract:

This study investigated the effects of three types of pre-reading activities (vocabulary definitions, text summary and pre-passage questions) on EFL learners’ English reading comprehension. On the basis of the results of a placement test administered to two hundred and thirty English students at Kerman Azad University, 200 subjects (one hundred intermediate and one hundred advanced) were selected.Four texts, two of them at intermediate level and two of them at advanced level were chosen. The data gathered was subjected to the statistical procedures of ANOVA. A close examination of the results through Tukey’s HSD showed the fact that the experimental groups performed better than the control group, highlighting the effect of the treatment on them. Also, the experimental group C (text summary), performed remarkably better than the other three groups (both experimental & control). Group B subjects, vocabulary definitions, performed better than groups A and D. The pre-passage questions group’s (D) performance showed higher scores than the control condition.

Keywords: pre-reading activities, text summary, vocabulary definition, and pre-passage questions, reading comprehension

Procedia PDF Downloads 304
2708 Stereo Motion Tracking

Authors: Yudhajit Datta, Hamsi Iyer, Jonathan Bandi, Ankit Sethia

Abstract:

Motion Tracking and Stereo Vision are complicated, albeit well-understood problems in computer vision. Existing softwares that combine the two approaches to perform stereo motion tracking typically employ complicated and computationally expensive procedures. The purpose of this study is to create a simple and effective solution capable of combining the two approaches. The study aims to explore a strategy to combine the two techniques of two-dimensional motion tracking using Kalman Filter; and depth detection of object using Stereo Vision. In conventional approaches objects in the scene of interest are observed using a single camera. However for Stereo Motion Tracking; the scene of interest is observed using video feeds from two calibrated cameras. Using two simultaneous measurements from the two cameras a calculation for the depth of the object from the plane containing the cameras is made. The approach attempts to capture the entire three-dimensional spatial information of each object at the scene and represent it through a software estimator object. In discrete intervals, the estimator tracks object motion in the plane parallel to plane containing cameras and updates the perpendicular distance value of the object from the plane containing the cameras as depth. The ability to efficiently track the motion of objects in three-dimensional space using a simplified approach could prove to be an indispensable tool in a variety of surveillance scenarios. The approach may find application from high security surveillance scenes such as premises of bank vaults, prisons or other detention facilities; to low cost applications in supermarkets and car parking lots.

Keywords: kalman filter, stereo vision, motion tracking, matlab, object tracking, camera calibration, computer vision system toolbox

Procedia PDF Downloads 288
2707 Entropy in a Field of Emergence in an Aspect of Linguo-Culture

Authors: Nurvadi Albekov

Abstract:

Communicative situation is a basis, which designates potential models of ‘constructed forms’, a motivated basis of a text, for a text can be assumed as a product of the communicative situation. It is within the field of emergence the models of text, that can be potentially prognosticated in a certain communicative situation, are designated. Every text can be assumed as conceptual system structured on the base of certain communicative situation. However in the process of ‘structuring’ of a certain model of ‘conceptual system’ consciousness of a recipient is able act only within the border of the field of emergence for going out of this border indicates misunderstanding of the communicative situation. On the base of communicative situation we can witness the increment of meaning where the synergizing of the informative model of communication, formed by using of the invariant units of a language system, is a result of verbalization of the communicative situation. The potential of the models of a text, prognosticated within the field of emergence, also depends on the communicative situation. The conception ‘the field of emergence’ is interpreted as a unit of the language system, having poly-directed universal structure, implying the presence of the core, the center and the periphery, including different levels of means of a functioning system of language, both in terms of linguistic resources, and in terms of extra linguistic factors interaction of which results increment of a text. The conception ‘field of emergence’ is considered as the most promising in the analysis of texts: oral, written, printed and electronic. As a unit of the language system field of emergence has several properties that predict its use during the study of a text in different levels. This work is an attempt analysis of entropy in a text in the aspect of lingua-cultural code, prognosticated within the model of the field of emergence. The article describes the problem of entropy in the field of emergence, caused by influence of the extra-linguistic factors. The increasing of entropy is caused not only by the fact of intrusion of the language resources but by influence of the alien culture in a whole, and by appearance of non-typical for this very culture symbols in the field of emergence. The borrowing of alien lingua-cultural symbols into the lingua-culture of the author is a reason of increasing the entropy when constructing a text both in meaning and in structuring level. It is nothing but artificial formatting of lexical units that violate stylistic unity of a phrase. It is marked that one of the important characteristics descending the entropy in the field of emergence is a typical similarity of lexical and semantic resources of the different lingua-cultures in aspects of extra linguistic factors.

Keywords: communicative situation, field of emergence, lingua-culture, entropy

Procedia PDF Downloads 325
2706 Understanding the Qualitative Nature of Product Reviews by Integrating Text Processing Algorithm and Usability Feature Extraction

Authors: Cherry Yieng Siang Ling, Joong Hee Lee, Myung Hwan Yun

Abstract:

The quality of a product to be usable has become the basic requirement in consumer’s perspective while failing the requirement ends up the customer from not using the product. Identifying usability issues from analyzing quantitative and qualitative data collected from usability testing and evaluation activities aids in the process of product design, yet the lack of studies and researches regarding analysis methodologies in qualitative text data of usability field inhibits the potential of these data for more useful applications. While the possibility of analyzing qualitative text data found with the rapid development of data analysis studies such as natural language processing field in understanding human language in computer, and machine learning field in providing predictive model and clustering tool. Therefore, this research aims to study the application capability of text processing algorithm in analysis of qualitative text data collected from usability activities. This research utilized datasets collected from LG neckband headset usability experiment in which the datasets consist of headset survey text data, subject’s data and product physical data. In the analysis procedure, which integrated with the text-processing algorithm, the process includes training of comments onto vector space, labeling them with the subject and product physical feature data, and clustering to validate the result of comment vector clustering. The result shows 'volume and music control button' as the usability feature that matches best with the cluster of comment vectors where centroid comments of a cluster emphasized more on button positions, while centroid comments of the other cluster emphasized more on button interface issues. When volume and music control buttons are designed separately, the participant experienced less confusion, and thus, the comments mentioned only about the buttons' positions. While in the situation where the volume and music control buttons are designed as a single button, the participants experienced interface issues regarding the buttons such as operating methods of functions and confusion of functions' buttons. The relevance of the cluster centroid comments with the extracted feature explained the capability of text processing algorithms in analyzing qualitative text data from usability testing and evaluations.

Keywords: usability, qualitative data, text-processing algorithm, natural language processing

Procedia PDF Downloads 242
2705 Effective Stacking of Deep Neural Models for Automated Object Recognition in Retail Stores

Authors: Ankit Sinha, Soham Banerjee, Pratik Chattopadhyay

Abstract:

Automated product recognition in retail stores is an important real-world application in the domain of Computer Vision and Pattern Recognition. In this paper, we consider the problem of automatically identifying the classes of the products placed on racks in retail stores from an image of the rack and information about the query/product images. We improve upon the existing approaches in terms of effectiveness and memory requirement by developing a two-stage object detection and recognition pipeline comprising of a Faster-RCNN-based object localizer that detects the object regions in the rack image and a ResNet-18-based image encoder that classifies the detected regions into the appropriate classes. Each of the models is fine-tuned using appropriate data sets for better prediction and data augmentation is performed on each query image to prepare an extensive gallery set for fine-tuning the ResNet-18-based product recognition model. This encoder is trained using a triplet loss function following the strategy of online-hard-negative-mining for improved prediction. The proposed models are lightweight and can be connected in an end-to-end manner during deployment to automatically identify each product object placed in a rack image. Extensive experiments using Grozi-32k and GP-180 data sets verify the effectiveness of the proposed model.

Keywords: retail stores, faster-RCNN, object localization, ResNet-18, triplet loss, data augmentation, product recognition

Procedia PDF Downloads 106