Search results for: labeled corpus

Authors: Lilia M. Fernando, Matthew K. Vasher, Evangelyn C. Alocilja

Abstract:

This paper describes the development of a DNA-based nanobiosensor to detect the dengue virus in mosquito using electrically active magnetic (EAM) nanoparticles as the concentrator and electrochemical transducer. The biosensor detection encompasses two sets of oligonucleotide probes that are specific to the dengue virus: the detector probe labeled with the EAM nanoparticles and the biotinylated capture probe. The DNA targets are double hybridized to the detector and the capture probes and concentrated from nonspecific DNA fragments by applying a magnetic field. Subsequently, the DNA sandwiched targets (EAM-detector probe–DNA target–capture probe-biotin) are captured on streptavidin modified screen printed carbon electrodes through the biotinylated capture probes. Detection is achieved electrochemically by measuring the oxidation–reduction signal of the EAM nanoparticles. Results indicate that the biosensor is able to detect the redox signal of the EAM nanoparticles at dengue DNA concentrations as low as 10 ng/ul.

Keywords: dengue, magnetic nanoparticles, mosquito, nanobiosensor

Procedia PDF Downloads 350

Authors: Ouazoul Abdelouahd, Jemjami Nadia

Abstract:

The concept of well-being at work is one of today's significant challenges in maintaining quality of life and managing psycho-social risks at work. Indeed, work in the sports sector has evolved, and this exponential evolution, marked by increasing demands and psychological, physical, and social challenges, which sometimes exceed the resources of sports actors, influences their sense of well-being at work. Well-being and burnout as antagonists provide information on the quality of working life in sports. The Basic aim of this literature review is to analyze the scientific corpus dealing with the subject of well-being at work in the sports sector while exploring the link between sports burnout and well-being. The results reveal the richness of the conceptual approaches and the difficulties of implementing them. Prospects for future research have, therefore, been put forward.

Keywords: Well-being, quality of life, Burnout, ; psycho-social risk, Sports sector

Procedia PDF Downloads 66

Authors: Ali Ulvi Uzer

Abstract:

Data reduction is an important topic in the field of pattern recognition applications. The basic concept is the reduction of multitudinous amounts of data down to the meaningful parts. The Principal Component Analysis (PCA) method is frequently used for data reduction. The Support Vector Machine (SVM) method is a discriminative classifier formally defined by a separating hyperplane. In other words, given labeled training data, the algorithm outputs an optimal hyperplane which categorizes new examples. This study offers a hybrid approach that uses the PCA for data reduction and Support Vector Machines (SVM) for classification. In order to detect the accuracy of the suggested system, two boreholes taken from the soil sample was used. The classification accuracies for this dataset were obtained through using ten-fold cross-validation method. As the results suggest, this system, which is performed through size reduction, is a feasible system for faster recognition of dataset so our study result appears to be very promising.

Keywords: feature selection, sequential forward selection, support vector machines, soil sample

Procedia PDF Downloads 440

Authors: Saba Rahimi, Ozan Oktay, Javier Alvarez-Valle, Sujeeth Bharadwaj

Abstract:

Successful application of deep learning in medical image analysis necessitates unprecedented amounts of labeled training data. Unlike conventional 2D applications, radiological images can be three-dimensional (e.g., CT, MRI), consisting of many instances within each image. The problem is exacerbated when expert annotations are required for effective pixel-wise labeling, which incurs exorbitant labeling effort and cost. Active learning is an established research domain that aims to reduce labeling workload by prioritizing a subset of informative unlabeled examples to annotate. Our contribution is a cost-effective approach for U-Net 3D models that uses Monte Carlo sampling to analyze pixel-wise uncertainty. Experiments on the AAPM 2017 lung CT segmentation challenge dataset show that our proposed framework can achieve promising segmentation results by using only 42% of the training data.

Keywords: image segmentation, active learning, convolutional neural network, 3D U-Net

Procedia PDF Downloads 135

Authors: Ouazoul Abdeloauhd, Jemjami Nadia

Abstract:

The concept of well-being at work is one of today's significant challenges in maintaining quality of life and managing psycho-social risks at work. Indeed, work in the sports sector has evolved, and this exponential evolution, marked by increasing demands and psychological, physical, and social challenges, which sometimes exceed the resources of sports actors, influences their sense of well-being at work. Well-being and burnout as antagonists provide information on the quality of working life in sports. The Basic aim of this literature review is to analyze the scientific corpus dealing with the subject of well-being at work in the sports sector while exploring the link between sports burnout and well-being. The results reveal the richness of the conceptual approaches and the difficulties of implementing them. Prospects for future research have, therefore, been put forward.

Keywords: well-being, burnout, quality of life, psycho-social risk, work on sports sector

Procedia PDF Downloads 75

Authors: Maha Wiss, Wael Khreich

Abstract:

Animals, objects, foods, plants, and other non-human terms are commonly used as a source of metaphors to describe females in formal and slang language. Comparing women to non-human items not only reflects cultural views that might conceptualize women as subordinates or in a lower position than humans, yet it conveys this degradation to the listeners. Moreover, the dehumanizing representation of females in the language normalizes the derogation and even encourages sexism and aggressiveness against women. Although dehumanization has been a popular research topic for decades, according to our knowledge, no studies have linked women's dehumanizing language to the machine learning field. Therefore, we introduce our research work as one of the first attempts to create a tool for the automated detection of the dehumanizing depiction of females in English texts. We also present the first labeled dataset on the charted topic, which is used for training supervised machine learning algorithms to build an accurate classification model. The importance of this work is that it accomplishes the first step toward mitigating dehumanizing language against females.

Keywords: gender bias, machine learning, NLP, women dehumanization

Procedia PDF Downloads 68

Authors: Shuo Liu

Abstract:

In Chinese text classification tasks, redundant words and phrases can interfere with the formation of extracted and analyzed text information, leading to a decrease in the accuracy of the classification model. To reduce irrelevant elements, extract and utilize text content information more efficiently and improve the accuracy of text classification models. In this paper, the text in the corpus is first extracted using the TextRank algorithm for abstraction, the words in the abstract are used as nodes to construct a text graph, and then the graph attention network (GAT) is used to complete the task of classifying the text. Testing on a Chinese dataset from the network, the classification accuracy was improved over the direct method of generating graph structures using text.

Keywords: Chinese natural language processing, text classification, abstract extraction, graph attention network

Procedia PDF Downloads 85

Authors: Karin Johansson, Ola Johansson

Abstract:

Recycling is an important aspect of a sustainable society and depends to a large extent on consumers’ willingness to provide the voluntary work needed to take the first critical step in many return logistics systems. Based on a systematic review of articles on recycling behavior, this paper presents and discusses the findings in relation to Fogg’s Behavioral Model (FBM). Through the analysis of a corpus of 72 articles, the most important research contributions on recycling behavior are summarized and discussed. The choice of using FBM as a framework provides a new way of viewing previous research findings, and aids in identifying knowledge gaps. Based on the review, this work identifies and discusses four areas of potential interest for future research.

Keywords: recycling, reverse logistics, solid waste management, sustainability

Procedia PDF Downloads 130

Authors: Steven Woods

Abstract:

The use of Corporate Social Responsibility Statements has become ubiquitous in society. The appeal to consumers by being a well-behaved social entity has become a strategy not just to ensure brand loyalty but also to further larger scale projects of corporate interests. Specifically, the use of CSR to position corporations as good planetary citizens involves not just self-promotion but also a way of transferring responsibility from systems to individuals. By using techniques labeled as “greenwashing” and emphasizing ethical consumption choices as the solution, corporations present themselves as good members of the community and pursuing sustainability. Ultimately, the primary function of Corporate Social Responsibility statements is to maintain the economic status quo of ongoing growth and consumption while presenting and environmentally progressive image to the public, as well as reassuring them corporate behavior is superior to government intervention. By analyzing the communication techniques utilized through content analysis of specific examples, along with an analysis of the frames of meaning constructed in the CSR statements, the practices of Corporate Responsibility and Sustainability will be addressed from an ethical perspective.

Keywords: corporate social responsibility, ethics, greenwashing, sustainability

Procedia PDF Downloads 60

Authors: Sefa Aksu, Ünal Kızıl

Abstract:

For this study, a town based soil database created in Gümüşçay District of Biga Town, Çanakkale, Turkey. Crop and livestock production are major activities in the district. Nutrient management is mainly based on commercial fertilizer application ignoring the livestock manure. Within the boundaries of district, 122 soil sampling points determined over the satellite image. Soil samples collected from the determined points with the help of handheld Global Positioning System. Labeled samples were sent to a commercial laboratory to determine 11 soil parameters including salinity, pH, lime, organic matter, nitrogen, phosphorus, potassium, iron, manganese, copper and zinc. Based on the test results soil maps for mentioned parameters were developed using remote sensing, GIS, and geostatistical analysis. In this study we developed a GIS database that will be used for soil nutrient management. Methods were explained and soil maps and their interpretations were summarized in the study.

Keywords: geostatistics, GIS, nutrient management, soil mapping

Procedia PDF Downloads 357

Authors: Takanori Tanaka, Daisuke Kitao, Daisuke Ikeda

Abstract:

The aurora is an attractive phenomenon but it is difficult to understand the whole mechanism of it. An approach of data-intensive science might be an effective approach to elucidate such a difficult phenomenon. To do that we need labeled data, which shows when and what types of auroras, have appeared. In this paper, we propose an image retrieval system for auroral all-sky images, some of which include discrete and diffuse aurora, and the other do not any aurora. The proposed system retrieves images which are similar to the query image by using a popular image recognition method. Using 300 all-sky images obtained at Tromso Norway, we evaluate two methods of image recognition methods with or without our original color filtering method. The best performance is achieved when SIFT with the color filtering is used and its accuracy is 81.7% for discrete auroras and 86.7% for diffuse auroras.

Keywords: data-intensive science, image classification, content-based image retrieval, aurora

Procedia PDF Downloads 439

Authors: Mostafa ‘Neo’ Mohsenvand, Mohammad Rasool Izadi, Pattie Maes

Abstract:

EEG-based biometrics (user identification) has been explored on small datasets of no more than 157 subjects. Here we show that the accuracy of modern supervised methods falls rapidly as the number of users increases to a few thousand. Moreover, supervised methods require a large amount of labeled data for training which limits their applications in real-world scenarios where acquiring data for training should not take more than a few minutes. We show that using contrastive learning for pre-training, it is possible to maintain high accuracy on a dataset of 2130 subjects while only using a fraction of labels. We compare 5 different self-supervised tasks for pre-training of the encoder where our proposed method achieves the accuracy of 96.4%, improving the baseline supervised models by 22.75% and the competing self-supervised model by 3.93%. We also study the effects of the length of the signal and the number of channels on the accuracy of the user-identification models. Our results reveal that signals from temporal and frontal channels contain more identifying features compared to other channels.

Keywords: brainprint, contrastive learning, electroencephalo-gram, self-supervised learning, user identification

Procedia PDF Downloads 146

Authors: Khalak Bin Khair, Saqib Jahir, Mohammed Ibrahim, Fahad Bin, Debajyoti Karmaker

Abstract:

Object detection is a computer technology that deals with searching through digital images and videos for occurrences of semantic elements of a particular class. It is associated with image processing and computer vision. On top of object detection, we detect camouflage objects within an image using Deep Learning techniques. Deep learning may be a subset of machine learning that's essentially a three-layer neural network Over 6500 images that possess camouflage properties are gathered from various internet sources and divided into 4 categories to compare the result. Those images are labeled and then trained and tested using vgg16 architecture on the jupyter notebook using the TensorFlow platform. The architecture is further customized using Transfer Learning. Methods for transferring information from one or more of these source tasks to increase learning in a related target task are created through transfer learning. The purpose of this transfer of learning methodologies is to aid in the evolution of machine learning to the point where it is as efficient as human learning.

Keywords: deep learning, transfer learning, TensorFlow, camouflage, object detection, architecture, accuracy, model, VGG16

Procedia PDF Downloads 125

Authors: Oublal Ali

Abstract:

Due scholarly attention given to Ludwig Wittgenstein since the appearance of Tractatus is resulted from revolutionary shift he has made in the conception of language. True, his first and foremost concern was to solve the issue of language philosophers failed to recognize. Not only Tracturain’s approach to language that argues for philosophers failure of understanding the logic of language, but also his later conception which is developed in philosophical investigations and the reminder of all his remarks. On such a basis, it is claimed that Wittgenstein’s theory of language should not be confined to the language within philosophical streams with this premise we therefore propose to analytically read one of the literary propositions in the sympathizer as linguistic corpus. Our investigation of the literary proposition weaves us into claiming that Wittgenstein’s language games -later philosophy- is apposite to the analysis of literary works thanks to the shift Wittgenstein has made from demarcated use of language to the multiplicity and non-uniformity of its use.

Keywords: language, context, use, language games, literary propositions

Procedia PDF Downloads 112

Authors: Aicha Aggoune, Abdelkrim Bouramoul, Mohamed Khiereddine Kholladi

Abstract:

Recent research shows that user profile as important element can improve heterogeneous information retrieval with its content. In this context, we present our indexing model for heterogeneous multimedia documents. This model is based on the combination of user profile to the indexing process. The general idea of our proposal is to operate the common concepts between the representation of a document and the definition of a user through his profile. These two elements will be added as additional indexing entities to enrich the heterogeneous corpus documents indexes. We have developed IRONTO domain ontology allowing annotation of documents. We will present also the developed tool validating the proposed model.

Keywords: indexing model, user profile, multimedia document, heterogeneous of sources, ontology

Procedia PDF Downloads 334

Authors: M. C. Padma, R. J. Prathibha

Abstract:

Parts of speech tagging is the process of assigning appropriate parts of speech tags to the words in a given text. The critical or crucial information needed for tagging a word come from its internal structure rather from its neighboring words. The internal structure of a word comprises of its morphological features and grammatical information. This paper presents a morpheme based parts of speech tagger for Kannada language. This proposed work uses hierarchical tag set for assigning tags. The system is tested on some Kannada words taken from EMILLE corpus. Experimental result shows that the performance of the proposed system is above 90%.

Keywords: hierarchical tag set, morphological analyzer, natural language processing, paradigms, parts of speech

Procedia PDF Downloads 280

Authors: Bruno C. Schimming, Leandro L. Martins, PatríCia F. F. Pinheiro, Raquel F. Domeniconi, FabríCio S. Oliveira

Abstract:

The agouti is a wildlife rodent that can be used as an alternative source of animal protein and this species has been raised in captivity in Brazil with the aim of providing meat. Thus, the knowledge of their reproductive biology and morphology of the reproductive organs is important. The objective of this study was to describe the morphology of epididymis in the Azara’s agouti, by light microscopy. Samples of epididymis were obtained from five adult Azara’s agouti (Dasyprocta azarae) during castration surgery performed at the Municipal Zoo of Catanduva, Brazil. Fragments of the epididymal regions (initial segment, caput, corpus and cauda) were collected. The biological samples were immediately fixed in paraformaldehyde for 24 hours, followed by histologic procedures comprising embedding in ParaplastTM (Sigma, St. Louis, MO, USA), sections of 5 µm, and staining with HE and Masson’s trichrome. The epididymis was a highly convoluted tubule that links the testis to the vas deferens. The epithelium lining was pseudostratified columnar surrounded by a periductal stroma. The epithelium contains several cell types: principal, basal, apical, clear, and hallo cells. Principal cells were the most abundant cell type. There were observed also migratory cells named halo cells. The caput epididymis was divided into two different regions: initial segment and caput. The initial segment has a very wide lumen, a high epithelium with conspicuous microvilli and the lumen was wide with exfoliated material. The other region of the caput epididymis, showed a lower epithelium when compared with the initial segment, large amounts of spermatozoa in the lumen, and a cytoplasmic vacuolization. This region presented many narrows cells. Many spermatozoa appeared in the lumen of corpus epididymis. The cauda region had a lower epithelium than the other epididymal regions in the agouti. The cauda epithelium presented plicae protruding into the lumen. Large amounts of spermatozoa are also present in the lumen. Small microvilli uniformly arranged so as to form a kind of “brush border” are observed on the apical surface of the cauda epithelium. The pattern of the epithelium lining the duct of the agouti epididymis does not differ greatly from that reported to other mammals, such as domestic and wildlife animals. These findings can cooperate with future investigations especially those related to rational exploration of these animals. All experimental procedures were approved by the institutional ethics committee (CEUA 796/2015). This study was supported by FAPESP (Grants 2015/23822-1).

Keywords: wildlife, testis excurrent ducts, epididymis, morphology

Procedia PDF Downloads 220

Authors: Attila Simon

Abstract:

The subject of this research is an element of Mithras' iconography. It is a new element in the series of research begun with the study of the Bull in the Boat motif. The stylized altars represented by the seven adjacent rectangles appear on only a small group of Mithraic reliefs, which may explain why they have received less attention and fewer attempts at decipherment than other motifs. Using Vermaseren's database, CIMRM (Corpus Inscriptionum et Monumentorum Religionis Mithriacae), one collected all the cases containing the motif under investigation, created a database of them grouped by location, then used a comparative method to compare the different forms of the motif and to isolate these cases, and finally evaluated the results. The aim of this research is to interpret the iconographic element in question and attempt to determine its place of origin. The study may provide an interpretation of a Mithraic representation that, to the best of the author's knowledge, has not been explained so far, and the question may generate scientific discourses.

Keywords: roman history, religion, Mithras, iconography

Procedia PDF Downloads 76

Authors: Ouazoul Abdelouahd, Jemjami Nadia

Abstract:

The concept of well-being at work is one of today's major challenges in maintaining quality of life and managing psycho-social risks at work. Indeed, work in the sports sector has evolved over time, and this exponential evolution, marked by increasing demands and psychological, physical and/or social challenges, which sometimes exceed the resources of sports players, influences their sense of well-being at work. Well-being and burnout as antagonists provide information on the quality of working life in sports. The main objective of this literature review was to examine the scientific corpus dealing with the subject of psychological well-being at work in the sports sector and, more specifically, with sports coaches while exploring the link between sports burnout and well-being. The results reveal the richness of the conceptual approaches and the difficulties of putting them into practice. Prospects for future research were put forward.

Keywords: psychological well-being, burnout, quality of life, sports coaching

Procedia PDF Downloads 17

Authors: Yufen Qin

Abstract:

Text sentiment transfer is an important branch of text style transfer. The goal is to generate text with another sentiment attribute based on a text with a specific sentiment attribute while maintaining the content and semantic information unrelated to sentiment unchanged in the process. There are currently two main challenges in this field: no parallel corpus and text attribute entanglement. In response to the above problems, this paper proposed a novel solution: Mask-Prompt-Rerank. Use the method of masking the sentiment words and then using prompt regeneration to transfer the sentence sentiment. Experiments on two sentiment benchmark datasets and one formality transfer benchmark dataset show that this approach makes the performance of small pre-trained language models comparable to that of the most advanced large models, while consuming two orders of magnitude less computing and memory.

Keywords: language model, natural language processing, prompt, text sentiment transfer

Procedia PDF Downloads 63

Authors: Zainab Ajab Mohideen, Kiran Kaur, A. Basheer Ahamadhu, Noor Azlinda Wan Jan, Sukmawati Muhammad

Abstract:

The corpus in the digital library is to provide an overview and evidence from library automation that can be used to justify the needs of the digital library. This paper disperses the approach and implementation of the digital library as part of best practices by the Automation Division at Hamzah Sendut Library of the University Science Malaysia (USM). The implemented digital library model emphasizes on the entire library collections, technical perspective, and automation solution. This model served as a foundation for digital library services as part of information delivery in the USM digital library. The approach to digital library includes discussion on key factors, design, architecture, and pragmatic model that has been collected, captured, and identified during the implementation stages. At present, the USM digital library has achieved the status of an Institutional Repository (IR).

Keywords: academic digital library, digital information system, digital library best practice, digital library model

Procedia PDF Downloads 538

Authors: Go Woon Kwak, Siyoung Jun, Soyun Maeng, Haeyoung Lee

Abstract:

Entry is an education platform used in South Korea, created to help students learn to program, in which they can learn to code while playing. Using the online version of the entry, teachers can easily assign programming homework to the student and the students can make programs simply by linking programming blocks. However, the programs may be made by others, so that the authors of the programs should be identified. In this paper, as the first step toward author identification of entry programs, we present an artificial neural network based classification approach to identify genders of authors of a program written in an entry. A neural network has been trained from labeled training data that we have collected. Our result in progress, although preliminary, shows that the proposed approach could be feasible to be applied to the online version of entry for gender identification of authors. As future work, we will first use a machine learning technique for age identification of entry programs, which would be the second step toward the author identification.

Keywords: artificial intelligence, author identification, deep neural network, gender identification, machine learning

Procedia PDF Downloads 305

Authors: Mohamad Sharif bin Mustaffa, Maizatul Akmam binti Abu Bakar, Mohd Harriszamani bin Abu Bakar

Abstract:

Most teens who gave birth to an illegitimate child will suffer from depression. This depression issues arising from the incident itself which contains the teen will feel ashamed because labeled as immoral, apart from that most teenagers go through this episode without support from their spouse or family. Teens also face the possibility of a large part with the will be born babies because the issue of too young and needs to go back to school. Teens also can not make a decision on the future of the baby later because they are too immature and no baby care skills other than financial issues, where young people themselves are still dependent on their family. This paper will look at how to apply neurofeedback can be used to see the level of depression experienced by teenagers who get pregnant out of wedlock. Play therapy that will help improve adolescent focus will be used for this purpose. Each level experienced by teenagers going through the phase of easy-to-high difficulty level. Apart from that a recovery module will also be developed as a whole to reduce the level of depression to enable the youth to perform routine healthy activities and can go back to school with cheerful feeling, motivated and active.

Keywords: neurofeedback, depression, pregnant, adolescent, illegitimate

Procedia PDF Downloads 423

Authors: Muhamad Afiq Faisal, Yahaya, Mohd Faizal, Hamzah

Abstract:

Genetic modification technology allows scientists to alter the genetic information of a particular organism. The technology allows the production of genetically modified organism (GMO) that has the enhanced property compared to the unmodified organism. The application of such technology is not only in agriculture industry, it is now has been applied extensively in biopharmaceutical industry such as transgenic vaccines. In Malaysia, Biosafety Act 2007 has been enacted in which all GMO-based products must be labeled with adequate information before being marketed. This paper aims to determine the awareness level amongst Malaysian consumers on the GM products available in the market and the efficiency of information supplied in the GM product labeling. The result of the survey will serve as a guideline for Malaysia government agency bodies to provide comprehensive yet efficient information to consumers for the purpose of GM product labeling in the near future. In conclusion, the efficiency of information delivery plays a vital role in ensuring that the information is being conveyed clearly to Malaysian consumers during the selection process of GM products available in the market.

Keywords: genetic modification technology, genetically modified organisms, genetically modified organism products labeling, Biosafety Act 2007

Procedia PDF Downloads 349

Authors: Isabelle Snyder, Travis Smith

Abstract:

Microgrid protection presents challenges to conventional protection techniques due to the low-induced fault current. Protection relays present in microgrid applications require a combination of settings groups to adjust based on the architecture of the microgrid in islanded and grid-connected modes. In a radial system where the microgrid is at the other end of the feeder, directional elements can be used to identify the direction of the fault current and switch settings groups accordingly (grid-connected or microgrid-connected). However, with multiple microgrid connections, this concept becomes more challenging, and the direction of the current alone is not sufficient to identify the source of the fault current contribution. ORNL has previously developed adaptive relaying schemes through other DOE-funded research projects that will be evaluated and used as a baseline for this research. The four protection techniques in this study are labeled as follows: (1) Adaptive Current only Protection System (ACPS), Intentional (2) Unbalanced Control for Protection Control (IUCPC), (3) Adaptive Protection System with Communication Controller (APSCC) (4) Adaptive Model-Driven Protective Relay (AMDPR).

Keywords: adaptive relaying, microgrid protection, sequence components, islanding detection

Procedia PDF Downloads 73

Authors: Umair Rashid

Abstract:

Normally softmax loss is used as the supervision signal in face recognition (FR) system, and it boosts the separability of features. In the last two years, a number of techniques have been proposed by reformulating the original softmax loss to enhance the discriminating power of Deep Convolutional Neural Networks (DCNNs) for FR system. To learn angularly discriminative features Cosine-Margin based softmax has been adjusted as monotonically decreasing angular function, that is the main challenge for angular based softmax. On that issue, we propose monotonically decreasing element for Cosine-Margin based softmax and also, we discussed the effect of different monotonically decreasing parameters on angular Margin softmax for FR system. We train the model on publicly available dataset CASIA- WebFace via our proposed monotonically decreasing parameters for cosine function and the tests on YouTube Faces (YTF, Labeled Face in the Wild (LFW), VGGFace1 and VGGFace2 attain the state-of-the-art performance.

Keywords: deep convolutional neural networks, cosine margin face recognition, softmax loss, monotonically decreasing parameter

Procedia PDF Downloads 82

Authors: Jiangbo Shi, Zeyu Gao, Chen Li

Abstract:

Nuclei segmentation is a fundamental task in digital pathology analysis and can be automated by deep learning-based methods. However, the development of such an automated method requires a large amount of data with precisely annotated masks which is hard to obtain. Training with weakly labeled data is a popular solution for reducing the workload of annotation. In this paper, we propose a novel meta-learning-based nuclei segmentation method which follows the label correction paradigm to leverage data with noisy masks. Specifically, we design a fully conventional meta-model that can correct noisy masks by using a small amount of clean meta-data. Then the corrected masks are used to supervise the training of the segmentation model. Meanwhile, a bi-level optimization method is adopted to alternately update the parameters of the main segmentation model and the meta-model. Extensive experimental results on two nuclear segmentation datasets show that our method achieves the state-of-the-art result. In particular, in some noise scenarios, it even exceeds the performance of training on supervised data.

Keywords: deep learning, histopathological image, meta-learning, nuclei segmentation, weak annotations

Procedia PDF Downloads 128

Authors: Li Qingjian, Li Ke, He Chun, Huang Yong

Abstract:

In this paper, the method of combining the Pohl Seidman's deep belief network with the self-organizing neural network is proposed to classify the target. This method is mainly aimed at the high nonlinearity of the hyperspectral image, the high sample dimension and the difficulty in designing the classifier. The main feature of original data is extracted by deep belief network. In the process of extracting features, adding known labels samples to fine tune the network, enriching the main characteristics. Then, the extracted feature vectors are classified into the self-organizing neural network. This method can effectively reduce the dimensions of data in the spectrum dimension in the preservation of large amounts of raw data information, to solve the traditional clustering and the long training time when labeled samples less deep learning algorithm for training problems, improve the classification accuracy and robustness. Through the data simulation, the results show that the proposed network structure can get a higher classification precision in the case of a small number of known label samples.

Keywords: DBN, SOM, pattern classification, hyperspectral, data compression

Procedia PDF Downloads 329

Authors: Dan Xie

Abstract:

The use of discourse markers (DMs) can play a crucial role in representing discourse interaction and pragmatic competence. Learners’ use of DMs and differences between native speakers (NSs) and non-native speakers (NNSs) in the use of various DMs have been the focus of considerable research attention. However, some commonly used DMs, such as you know, have not received as much attention in comparative studies, especially in the Chinese context. This study analyses data in two corpora (COLSEC and Spoken BNC 2014 (14-25)) to investigate how Chinese learners differ from NNSs in their use of the DM you know and its functions in speech. The results show that there is a significant difference between the two corpora in terms of the frequency of use of you know. In terms of the functions of you know, the study shows that six functions can all be present in both corpora, although there are significant differences between the five functional dimensions, especially in introducing a claim linked to the prior discourse and highlighting particular points in the discourse. It is hoped to show empirically how Chinese learners and NSs use DMs differently.

Keywords: you know, discourse marker, native speaker, Chinese learner

Procedia PDF Downloads 60

Authors: Gorkem Algan, Ilkay Ulusoy, Saban Gonul, Banu Turgut, Berker Bakbak

Abstract:

Computer vision systems recently made a big leap thanks to deep neural networks. However, these systems require correctly labeled large datasets in order to be trained properly, which is very difficult to obtain for medical applications. Two main reasons for label noise in medical applications are the high complexity of the data and conflicting opinions of experts. Moreover, medical imaging datasets are commonly tiny, which makes each data very important in learning. As a result, if not handled properly, label noise significantly degrades the performance. Therefore, a label-noise-robust learning algorithm that makes use of the meta-learning paradigm is proposed in this article. The proposed solution is tested on retinopathy of prematurity (ROP) dataset with a very high label noise of 68%. Results show that the proposed algorithm significantly improves the classification algorithm's performance in the presence of noisy labels.

Keywords: deep learning, label noise, robust learning, meta-learning, retinopathy of prematurity

Procedia PDF Downloads 145