Search results for: speech dataset

Authors: Vijaya K. Narne, Gerard Loquet, Tobias Piechowiak, Dorte Hammershoi, Jesper H. Schmidt

Abstract:

BEAR, which stands for better hearing rehabilitation is a large-scale project in Denmark designed and executed by three national universities, three hospitals, and the hearing aid industry with the aim to improve hearing aid fitting. A total of 1963 hearing impaired people were included and were segmented into subgroups based on hearing-loss, demographics, audiological and questionnaires data (i.e., the speech, spatial and qualities of hearing scale [SSQ-12] and the International Outcome Inventory for Hearing-Aids [IOI-HA]). With the aim to provide a better hearing-aid fit to individual patients, we applied modern machine learning techniques with traditional audiograms rule-based systems. Results show that age, speech discrimination scores, and audiogram configurations were evolved as important parameters in characterizing sub-population from the data-set. The attempt to characterize sub-population reveal a clearer picture about the individual hearing difficulties encountered and the benefits derived from more individualized hearing aids.

Keywords: hearing loss, audiological data, machine learning, hearing aids

Procedia PDF Downloads 141

Authors: Wassim Katerji, Mercedes Farjas, Carmen Morillo

Abstract:

Terrain-Based Analysis results in derived products from an input DEM and these products are needed to perform various analyses. To efficiently use these products in decision-making, their accuracies must be estimated systematically. This paper proposes a procedure to assess the accuracy of these derived products, by calculating the accuracy of the slope dataset and its significance, taking as an input the accuracy of the DEM. Based on the output of previously published research on modeling the relative accuracy of a DEM, specifically ASTER and SRTM DEMs with Lebanon coverage as the area of study, analysis have showed that ASTER has a low significance in the majority of the area where only 2% of the modeled terrain has 50% or more significance. On the other hand, SRTM showed a better significance, where 37% of the modeled terrain has 50% or more significance. Statistical analysis deduced that the accuracy of the slope dataset, calculated on a cell-by-cell basis, is highly correlated to the accuracy of the input DEM. However, this correlation becomes lower between the slope accuracy and the slope significance, whereas it becomes much higher between the modeled slope and the slope significance.

Keywords: terrain-based analysis, slope, accuracy assessment, Digital Elevation Model (DEM)

Procedia PDF Downloads 434

Authors: Mohamed Mounir Tellache, Hiroyuki Kambara, Yasuharu Koike, Makoto Miyakoshi, Natsue Yoshimura

Abstract:

This study explores the practicality of using electroencephalographic (EEG) independent components to predict eight-direction finger movements in pseudo-real-time. Six healthy participants with individual-head MRI images performed finger movements in eight directions with two different arm configurations. The analysis was performed in two stages. The first stage consisted of using independent component analysis (ICA) to separate the signals representing brain activity from non-brain activity signals and to obtain the unmixing matrix. The resulting independent components (ICs) were checked, and those reflecting brain-activity were selected. Finally, the time series of the selected ICs were used to predict eight finger-movement directions using Sparse Logistic Regression (SLR). The second stage consisted of using the previously obtained unmixing matrix, the selected ICs, and the model obtained by applying SLR to classify a different EEG dataset. This method was applied to two different settings, namely the single-participant level and the group-level. For the single-participant level, the EEG dataset used in the first stage and the EEG dataset used in the second stage originated from the same participant. For the group-level, the EEG datasets used in the first stage were constructed by temporally concatenating each combination without repetition of the EEG datasets of five participants out of six, whereas the EEG dataset used in the second stage originated from the remaining participants. The average test classification results across datasets (mean ± S.D.) were 38.62 ± 8.36% for the single-participant, which was significantly higher than the chance level (12.50 ± 0.01%), and 27.26 ± 4.39% for the group-level which was also significantly higher than the chance level (12.49% ± 0.01%). The classification accuracy within [–45°, 45°] of the true direction is 70.03 ± 8.14% for single-participant and 62.63 ± 6.07% for group-level which may be promising for some real-life applications. Clustering and contribution analyses further revealed the brain regions involved in finger movement and the temporal aspect of their contribution to the classification. These results showed the possibility of using the ICA-based method in combination with other methods to build a real-time system to control prostheses.

Keywords: brain-computer interface, electroencephalography, finger motion decoding, independent component analysis, pseudo real-time motion decoding

Procedia PDF Downloads 129

Authors: Syed Hamid Farooq Bukhari

Abstract:

This study intends to clarify the Friday sermon by evaluating the formation of its discourse, the composition, and selection of its subject matters, the structure, and functions of its rules as well as the outline of its communication proceeds, and the distinctiveness of its words along with definite provisions. In this research, a qualitative and descriptive method is used to draw out conclusions. This paper considers the sermon mechanism of the speech and advances it contextually. The information was composed in Pakistan and several of its mosques supposing the imams of the city and the location of the mosques. The presentation and analysis of the facts have directed to the subsequent conclusions: (1) the Friday sermon holds verbal discussion that has habitual and classic formation, (2) the approaches of the formation of the subjects consist of storytelling, quotation as well as the use of accepted terms, (3) the composition of the codes involves Arabic, English, Urdu, and many other local languages, (4) the expressions of the speech include all types of sermon acts, (5) different requisites emerge in the sermons demonstrating that the Friday sermon functions as an index or usage of verbal communication in an exacting field.

Keywords: Friday, sermons, Pakistan, social

Procedia PDF Downloads 155

Authors: Geovana Soncin, Larissa Berti

Abstract:

Some aspects of prosody acquisition remain still unclear, especially regarding atypical speech development processes. This work deals with prosody acquisition and its implications for clinical purposes. Therefore, we analyze speech samples produced by adult speakers, children in typical language development, and children with phonological disorders. Phonological disorder comprises deviating manifestations characterized by inconsistencies in the phonological representation of a linguistic system under acquisition. The clinical assessment is performed mostly based on contrasts whose manifestations occur in the segmental level of a phonological system. Prosodic organization of spoken utterances is not included in the standard assessment. However, assuming that prosody is part of the phonological system, it was hypothesized that children with Phonological Disorders could present inconsistencies that also occur at a prosodic level. Based on this hypothesis, the paper aims to analyze contrastive focus marking in the speech of children with Phonological Disorders in comparison with the speech of children under Typical Language Development and adults. The participants of all groups were native speakers of Brazilian Portuguese. The investigation was designed in such a way as to identify differences and similarities among the groups that could be interpreted as clues of normal or deviant processes of prosody acquisition. Contrastive focus in Brazilian Portuguese is marked by increasing duration, f0, and intensity on the focused element as well as by a particular type of pitch accent (L*+H). Thirty-nine subjects participated, thirteen from each group. Acoustic analysis was performed, considering duration, intensity, and intonation as parameters. Children with PD were recruited in sessions from a service provided by Speech-Language Pathology Therapy; children in TD, paired in age and sex with the first group, were recruited in a regular school; and 20-24 years old adults were recruited from a University class. In a game prepared to elicit focused sentences, all of them produced the sentence “Girls love red dress,” marking focus on different syntactic positions: subject, verb, and object. Results showed that adults, children in typical language development, and children with Phonological Disorders marked contrastive focus differently: typical children used all parameters like adults do; however, in comparison with them, they exaggerated duration and, in the opposite direction, they did not increase f0 in a sufficient magnitude as adults; children with Phonological Disorder presented inconsistencies in duration, not increasing it in some syntactic positions, and also in intonation, not producing the representative pitch accent of contrastive focus. The results suggest prosody is also affected by phonological disorder and give clues of developmental processes of prosody acquisition.

Keywords: Brazilian Portuguese, contrastive focus, phonological disorder, prosody acquisition

Procedia PDF Downloads 75

Authors: Simona Maria Ignat

Abstract:

Bullying online has been an increasing studied topic during the last years. Different approaches, psychological, linguistic, or computational, have been applied. To our best knowledge, a definition and a set of characteristics of phenomenon agreed internationally as a common framework are still waiting for answers. Thus, the objectives of this paper are the identification of bullying utterances on Twitter and their algorithms. This research paper is focused on the identification of words or groups of words, categorized as “utterances”, with bullying effect, from Twitter platform, extracted on a set of legislative criteria. This set is the result of analysis followed by synthesis of law documents on bullying(online) from United States of America, European Union, and Ireland. The outcome is a linguistic corpus with approximatively 10,000 entries. The methods applied to the first objective have been the following. The discourse analysis has been applied in identification of keywords with bullying effect in texts from Google search engine, Images link. Transcription and anonymization have been applied on texts grouped in CL1 (Corpus linguistics 1). The keywords search method and the legislative criteria have been used for identifying bullying utterances from Twitter. The texts with at least 30 representations on Twitter have been grouped. They form the second corpus linguistics, Bullying utterances from Twitter (CL2). The entries have been identified by using the legislative criteria on the the BoW method principle. The BoW is a method of extracting words or group of words with same meaning in any context. The methods applied for reaching the second objective is the conversion of parts of speech to alphabetical and numerical symbols and writing the bullying utterances as algorithms. The converted form of parts of speech has been chosen on the criterion of relevance within bullying message. The inductive reasoning approach has been applied in sampling and identifying the algorithms. The results are groups with interchangeable elements. The outcomes convey two aspects of bullying: the form and the content or meaning. The form conveys the intentional intimidation against somebody, expressed at the level of texts by grammatical and lexical marks. This outcome has applicability in the forensic linguistics for establishing the intentionality of an action. Another outcome of form is a complex of graphemic variations essential in detecting harmful texts online. This research enriches the lexicon already known on the topic. The second aspect, the content, revealed the topics like threat, harassment, assault, or suicide. They are subcategories of a broader harmful content which is a constant concern for task forces and legislators at national and international levels. These topic – outcomes of the dataset are a valuable source of detection. The analysis of content revealed algorithms and lexicons which could be applied to other harmful contents. A third outcome of content are the conveyances of Stylistics, which is a rich source of discourse analysis of social media platforms. In conclusion, this corpus linguistics is structured on legislative criteria and could be used in various fields.

Keywords: corpus linguistics, cyberbullying, legislation, natural language processing, twitter

Procedia PDF Downloads 72

Authors: Molakala Mourya Vardhan Reddy, Kandimala Revanth, Koduru Sumanth, Beena B. M.

Abstract:

Emergency response and recovery operations are severely hampered by natural catastrophes, especially floods. Understanding post-flood scenarios is essential to disaster management because it facilitates quick evaluation and decision-making. To this end, we introduce FloodNet, a brand-new high-resolution aerial picture collection created especially for comprehending post-flood scenes. A varied collection of excellent aerial photos taken during and after flood occurrences make up FloodNet, which offers comprehensive representations of flooded landscapes, damaged infrastructure, and changed topographies. The dataset provides a thorough resource for training and assessing computer vision models designed to handle the complexity of post-flood scenarios, including a variety of environmental conditions and geographic regions. Pixel-level semantic segmentation masks are used to label the pictures in FloodNet, allowing for a more detailed examination of flood-related characteristics, including debris, water bodies, and damaged structures. Furthermore, temporal and positional metadata improve the dataset's usefulness for longitudinal research and spatiotemporal analysis. For activities like flood extent mapping, damage assessment, and infrastructure recovery projection, we provide baseline standards and evaluation metrics to promote research and development in the field of post-flood scene comprehension. By integrating FloodNet into machine learning pipelines, it will be easier to create reliable algorithms that will help politicians, urban planners, and first responders make choices both before and after floods. The goal of the FloodNet dataset is to support advances in computer vision, remote sensing, and disaster response technologies by providing a useful resource for researchers. FloodNet helps to create creative solutions for boosting communities' resilience in the face of natural catastrophes by tackling the particular problems presented by post-flood situations.

Keywords: image classification, segmentation, computer vision, nature disaster, unmanned arial vehicle(UAV), machine learning.

Procedia PDF Downloads 60

Authors: Azene Zenebe

Abstract:

Deep learning is a subset of machine learning which incorporates techniques for the construction of artificial neural networks and found to be useful for modeling complex problems with large dataset. Deep learning requires a very high power computational and longer time for training. By aggregating computing power, high performance computer (HPC) has emerged as an approach to resolving advanced problems and performing data-driven research activities. Cyber threat intelligence (CIT) is actionable information or insight an organization or individual uses to understand the threats that have, will, or are currently targeting the organization. Results of review of literature will be presented along with results of experimental study that compares the performance of tree-based and function-base machine learning including deep learning algorithms using secondary dataset collected from darknet.

Keywords: deep-learning, cyber security, cyber threat modeling, tree-based machine learning, function-based machine learning, data science

Procedia PDF Downloads 139

Authors: Maryam Zarabian, Hector Guzman, Pedro Pereira-Almao, Abraham Fapojuwo

Abstract:

Dry reforming of methane (DRM) has sparked significant industrial and scientific interest not only as a viable alternative for addressing the environmental concerns of two main contributors of the greenhouse effect, i.e., carbon dioxide (CO₂) and methane (CH₄), but also produces syngas, i.e., a mixture of hydrogen (H₂) and carbon monoxide (CO) utilized by a wide range of downstream processes as a feedstock for other chemical productions. In this study, we develop an AI-enable syngas production model to tackle the problem of achieving an equivalent H₂/CO ratio [1:1] with respect to the most efficient conversion. Firstly, the unsupervised density-based spatial clustering of applications with noise (DBSAN) algorithm removes outlier data points from the original experimental dataset. Then, random forest (RF) and deep neural network (DNN) models employ the error-free dataset to predict the DRM results. DNN models inherently would not be able to obtain accurate predictions without a huge dataset. To cope with this limitation, we employ reusing pre-trained layers’ approaches such as transfer learning and greedy layer-wise pretraining. Compared to the other deep models (i.e., pure deep model and transferred deep model), the greedy layer-wise pre-trained deep model provides the most accurate prediction as well as similar accuracy to the RF model with R² values 1.00, 0.999, 0.999, 0.999, 0.999, and 0.999 for the total outlet flow, H₂/CO ratio, H₂ yield, CO yield, CH₄ conversion, and CO₂ conversion outputs, respectively.

Keywords: artificial intelligence, dry reforming of methane, artificial neural network, deep learning, machine learning, transfer learning, greedy layer-wise pretraining

Procedia PDF Downloads 71

Authors: Ahlem Abid, Farah Jemili

Abstract:

Intrusion detection has been the subject of numerous studies in industry and academia, but cyber security analysts always want greater precision and global threat analysis to secure their systems in cyberspace. To improve intrusion detection system, the visualisation of the security events in form of graphs and diagrams is important to improve the accuracy of alerts. In this paper, we propose an approach of an IDS based on cloud computing, big data technique and using a machine learning graph algorithm which can detect in real time different attacks as early as possible. We use the MAWILab intrusion detection dataset . We choose Microsoft Azure as a unified cloud environment to load our dataset on. We implement the k2 algorithm which is a graphical machine learning algorithm to classify attacks. Our system showed a good performance due to the graphical machine learning algorithm and spark structured streaming engine.

Keywords: Apache Spark Streaming, Graph, Intrusion detection, k2 algorithm, Machine Learning, MAWILab, Microsoft Azure Cloud

Procedia PDF Downloads 133

Authors: Carla Lopes

Abstract:

It is clear nowadays that the proper performance of different speech acts is one of the most difficult obstacles that a foreign language learner has to overcome to be considered communicatively competent. This communication presents the results of an investigation on the pragmatic performance of Portuguese Language students at the University of Macau. The research discussed herein is based on a survey consisting of fourteen speaking situations to which the participants must respond in writing, and that includes different types of speech acts: apology, response to a compliment, refusal, complaint, disagreement and the understanding of the illocutionary force of indirect speech acts. The responses were classified in a five levels Likert scale (quantified from 1 to 5) according to their suitability for the particular situation. In general terms, we can summarize that about 45% of the respondents' answers were pragmatically competent, 10 % were acceptable and 45 % showed weaknesses at socio-pragmatic competence level. Given that the linguistic deviations were not taken into account, we can conclude that the faults are of cultural origin. It is natural that in the presence of orthogonal cultures, such as Chinese and Portuguese, there are failures of this type, barely solved in the four years of the undergraduate program. The target population, native speakers of Cantonese or Mandarin, make their first contact with the English language before joining the Bachelor of Portuguese Language. An analysis of the socio - pragmatic failures in the respondents’ answers suggests the conclusion that many of them are due to the lack of cultural knowledge. They try to compensate for this either using their native culture or resorting to a Western culture that they consider close to the Portuguese, that is the English or US culture, previously studied, and also widely present in the media and on the internet. This phenomenon, known as 'pragmatic transfer', can result in a linguistic behavior that may be considered inauthentic or pragmatically awkward. The resulting speech act is grammatically correct but is not pragmatically feasible, since it is not suitable to the culture of the target language, either because it does not exist or because the conditions of its use are in fact different. Analysis of the responses also supports the conclusion that these students present large deviations from the expected and stereotyped behavior of Chinese students. We can speculate while this linguistic behavior is the consequence of the Macao globalization that culturally casts the students, makes them more open, and distinguishes them from the typical Chinese students.

Keywords: Portuguese foreign language, pragmatic failures, pragmatic transfer, pragmatic competence

Procedia PDF Downloads 199

Authors: Chirag Kothawade, Gourie Jagtap, PreetKaur Relusinghani, Vedang Chavan, Smitha S. Bhosale

Abstract:

Psychological research has proven that music relieves stress, elevates mood, and is responsible for the release of “feel-good” chemicals like oxytocin, serotonin, and dopamine. It comes as no surprise that music has been a popular tool in rehabilitation centers and therapy for various disorders, thus with the interminably rising numbers of people facing mental health-related issues across the globe, addressing mental health concerns is more crucial than ever. Despite the existing music recommendation systems, there is a dearth of holistically curated algorithms that take care of the needs of users. Given that, an undeniable majority of people turn to music on a regular basis and that music has been proven to increase cognition, memory, and sleep quality while reducing anxiety, pain, and blood pressure, it is the need of the hour to fashion a product that extracts all the benefits of music in the most extensive and deployable method possible. Our project aims to ameliorate our users’ mental state by building a comprehensive mood-based music recommendation system called “Viby”.

Keywords: language, communication, speech recognition, interaction

Procedia PDF Downloads 47

Authors: Heramb Somthankar, Anujoy Chakraborty

Abstract:

This paper presents a classification study on yield prediction in semiconductor production using machine learning approaches. A complicated semiconductor production process is generally monitored continuously by signals acquired from sensors and measurement sites. A monitoring system contains a variety of signals, all of which contain useful information, irrelevant information, and noise. In the case of each signal being considered a feature, "Feature Selection" is used to find the most relevant signals. The open-source UCI SECOM Dataset provides 1567 such samples, out of which 104 fail in quality assurance. Feature extraction and selection are performed on the dataset, and useful signals were considered for further study. Afterward, common machine learning algorithms were employed to predict whether the signal yields pass or fail. The most relevant algorithm is selected for prediction based on the accuracy and loss of the ML model.

Keywords: deep learning, feature extraction, feature selection, machine learning classification algorithms, semiconductor production monitoring, signal processing, time-series analysis

Procedia PDF Downloads 98

Authors: Benjamin Bobbia, Matthias Picard

Abstract:

In recent years active learning methodologies based on the representativity of the data seems more promising to limit overfitting. The presented query methodology for regression using the Wasserstein distance measuring the representativity of our labelled dataset compared to the global distribution. In this work a crucial use of GroupSort Neural Networks is made therewith to draw a double advantage. The Wasserstein distance can be exactly expressed in terms of such neural networks. Moreover, one can provide explicit bounds for their size and depth together with rates of convergence. However, heterogeneity of the dataset is also considered by weighting the Wasserstein distance with the error of approximation at the previous step of active learning. Such an approach leads to a reduction of overfitting and high prediction performance after few steps of query. After having detailed the methodology and algorithm, an empirical study is presented in order to investigate the range of our hyperparameters. The performances of this method are compared, in terms of numbers of query needed, with other classical and recent query methods on several UCI datasets.

Keywords: active learning, Lipschitz regularization, neural networks, optimal transport, regression

Procedia PDF Downloads 70

Authors: Syed Asif Hassan, Syed Atif Hassan

Abstract:

Multidrug-resistant Tuberculosis (MDR-TB) is an infection caused by the resistant strains of Mycobacterium tuberculosis that do not respond either to isoniazid or rifampicin, which are the most important anti-TB drugs. The increase in the occurrence of a drug-resistance strain of MTB calls for an intensive search of novel target-based therapeutics. In this context LprG (Rv1411c) a lipoprotein from MTB plays a pivotal role in the immune evasion of Mtb leading to survival and propagation of the bacterium within the host cell. Therefore, a machine learning method will be developed for generating a computational model that could predict for a potential anti LprG activity of the novel antituberculosis compound. The present study will utilize dataset from PubChem database maintained by National Center for Biotechnology Information (NCBI). The dataset involves compounds screened against MTB were categorized as active and inactive based upon PubChem activity score. PowerMV, a molecular descriptor generator, and visualization tool will be used to generate the 2D molecular descriptors for the actives and inactive compounds present in the dataset. The 2D molecular descriptors generated from PowerMV will be used as features. We feed these features into three different classifiers, namely, random forest, a deep neural network, and a recurring neural network, to build separate predictive models and choosing the best performing model based on the accuracy of predicting novel antituberculosis compound with an anti LprG activity. Additionally, the efficacy of predicted active compounds will be screened using SMARTS filter to choose molecule with drug-like features.

Keywords: antituberculosis drug, classifier, machine learning, molecular descriptors, prediction

Procedia PDF Downloads 377

Authors: Made Rahayu Putri Saron, Mochamad Nizar Palefi Ma’ady

Abstract:

Communication skills are important in everyday life, communication can be done verbally in the form of oral or written and nonverbal in the form of expressions or body movements. Good communication should be able to provide information clearly, and there is feedback from the speaker and listener. However, it is often found that the information conveyed is not clear, and there is no feedback from the listeners, so it cannot be ensured that the communication is effective and understandable. The speaker's understanding of the topic is one of the supporting factors for the listener to be able to accept the meaning of the conversation. However, based on the results of the literature review, it found that the influence factors of person speaking style are as follows: (i) environmental conditions; (ii) voice, articulation, and accent; (iii) gender; (iv) personality; (v) speech disorders (Dysarthria); when speaking also have an important influence on speaker’s speaking style. It can be concluded the factors that support understanding and comfort of the listener are dependent on the nature of the speaker (environmental conditions, voice, gender, personality) or also it the speaker have speech disorders.

Keywords: listener, public speaking, speaking style, understanding, and comfortable factor

Procedia PDF Downloads 147

Authors: Peng Wang, Tigang Jiang

Abstract:

The computer based lip reading method of different languages cannot be universal. At present, for the research of Chinese lip reading, whether the work on data sets or recognition algorithms, is far from mature. In this paper, we study the Chinese lipreading method based on machine learning, and propose a Chinese Sentence-level lip-reading network (CNLipNet) model which consists of spatio-temporal convolutional neural network(CNN), recurrent neural network(RNN) and Connectionist Temporal Classification (CTC) loss function. This model can map variable-length sequence of video frames to Chinese Pinyin sequence and is trained end-to-end. More over, We create CNLRS, a Chinese Lipreading Dataset, which contains 5948 samples and can be shared through github. The evaluation of CNLipNet on this dataset yielded a 41% word correct rate and a 70.6% character correct rate. This evaluation result is far superior to the professional human lip readers, indicating that CNLipNet performs well in lipreading.

Keywords: lipreading, machine learning, spatio-temporal, convolutional neural network, recurrent neural network

Procedia PDF Downloads 112

Authors: Haonan Hu, Shuge Lei, Dasheng Sun, Huabin Zhang, Kehong Yuan, Jian Dai, Jijun Tang

Abstract:

This paper focuses on the classification task of breast ultrasound images and conducts research on the reliability measurement of classification results. A dual-channel evaluation framework was developed based on the proposed inference reliability and predictive reliability scores. For the inference reliability evaluation, human-aligned and doctor-agreed inference rationals based on the improved feature attribution algorithm SP-RISA are gracefully applied. Uncertainty quantification is used to evaluate the predictive reliability via the test time enhancement. The effectiveness of this reliability evaluation framework has been verified on the breast ultrasound clinical dataset YBUS, and its robustness is verified on the public dataset BUSI. The expected calibration errors on both datasets are significantly lower than traditional evaluation methods, which proves the effectiveness of the proposed reliability measurement.

Keywords: medical imaging, ultrasound imaging, XAI, uncertainty measurement, trustworthy AI

Procedia PDF Downloads 77

Authors: Xiaoxiao Li, Shuangcheng Jia, Qian Li

Abstract:

Keypoint detection has always been a challenge in the field of image recognition. This paper proposes a novelty keypoint detection method which is called Multi-Scale Feature Fusion Convolutional Network with Attention (MFFCNA). We verified that the multi-scale features with the attention mechanism module have better feature expression capability. The feature fusion between different scales makes the information that the network model can express more abundant, and the network is easier to converge. On our self-made street sign corner dataset, we validate the MFFCNA model with an accuracy of 97.8% and a recall of 81%, which are 5 and 8 percentage points higher than the HRNet network, respectively. On the COCO dataset, the AP is 71.9%, and the AR is 75.3%, which are 3 points and 2 points higher than HRNet, respectively. Extensive experiments show that our method has a remarkable improvement in the keypoint recognition tasks, and the recognition effect is better than the existing methods. Moreover, our method can be applied not only to keypoint detection but also to image classification and semantic segmentation with good generality.

Keywords: keypoint detection, feature fusion, attention, semantic segmentation

Procedia PDF Downloads 108

Authors: Nkechinyere Joy Olawuyi, Babajide Samuel Afolabi, Bola Ibitoye

Abstract:

This study collected a preprocessed dataset of chest radiographs and formulated a deep neural network model for detecting abnormalities. It also evaluated the performance of the formulated model and implemented a prototype of the formulated model. This was with the view to developing a deep neural network model to automatically classify abnormalities in chest radiographs. In order to achieve the overall purpose of this research, a large set of chest x-ray images were sourced for and collected from the CheXpert dataset, which is an online repository of annotated chest radiographs compiled by the Machine Learning Research Group, Stanford University. The chest radiographs were preprocessed into a format that can be fed into a deep neural network. The preprocessing techniques used were standardization and normalization. The classification problem was formulated as a multi-label binary classification model, which used convolutional neural network architecture to make a decision on whether an abnormality was present or not in the chest radiographs. The classification model was evaluated using specificity, sensitivity, and Area Under Curve (AUC) score as the parameter. A prototype of the classification model was implemented using Keras Open source deep learning framework in Python Programming Language. The AUC ROC curve of the model was able to classify Atelestasis, Support devices, Pleural effusion, Pneumonia, A normal CXR (no finding), Pneumothorax, and Consolidation. However, Lung opacity and Cardiomegaly had a probability of less than 0.5 and thus were classified as absent. Precision, recall, and F1 score values were 0.78; this implies that the number of False Positive and False Negative is the same, revealing some measure of label imbalance in the dataset. The study concluded that the developed model is sufficient to classify abnormalities present in chest radiographs into present or absent.

Keywords: transfer learning, convolutional neural network, radiograph, classification, multi-label

Procedia PDF Downloads 105

Authors: Mercy Itohan Aruya

Abstract:

This paper focuses on Usen, a speech form enclaved by the Edo communities in Ovia South West Local Government Area of Edo State, Nigeria. Usen lies at the border between Edo and the Osun state in Nigeria and has a population size of about a hundred and eighty thousand native speakers (2006 population census of Nigeria). Usen, as it is spoken today is highly endangered and it is serious struggling for survival. The aim, therefore, is to ascertain the linguistics status of Usen using a lexicostatical approach. Lexicostatics is a linguistic technique employed in accessing the degree of linguistic divergence or relatedness between two or more languages based on the proportion of cognates. Data for this study were collected from competent native speakers whose ages fall within the range of 40-65. The instrument for this study is the Ibadan 400 word-list of basic items which are collected with of a digital voice recorder. Our major finding in this paper reveals and establishes the facts that Usen speech form is not a dialect but a language of its own. However, Usen is more related to Yoruba than Edo as the degree of relatedness between Usen and Yoruba is 56.14% while that between Usen and Edo is about 21.4% as shown in this research effort.

Keywords: Usen, lexicostatistics, cognate words, language status

Procedia PDF Downloads 189

Authors: Chi-Wen Huang, Hung-Chi Chen

Abstract:

Total laryngectomy affects swallowing, speech functions and life quality in the head and neck cancer. Voice restoration plays an important role in social activities and communication. Several techniques have been developed for voice restoration and reported to improve the life quality. However, the rehabilitation program for voice reconstruction by using the ileocolon flap still unclear. A retrospective study was done, and the patients' data were drawn from the medical records between 2010 and 2016 who underwent voice reconstruction by ileocolon flap after laryngectomy. All of them were trained to swallow first; then, the voice rehabilitation was started. The outcome of voice was evaluated after 6 months using the 4-point scoring scale. In our result, 9.8% patients could give very clear voice so everyone could understand their speech, 61% patients could be understood well by families and friends, 20.2% patients could only talk with family, and 9% patients had difficulty to be understood. Moreover, the 57% patients did not need a second surgery, but in 43% patients voice was made clear by a second surgery. In this study, we demonstrated that the rehabilitation program after voice reconstruction with ileocolon flap for post-laryngectomy patients is important because the anatomical structure is different from the normal larynx.

Keywords: post-laryngectomy, ileocolon flap, rehabilitation, voice reconstruction

Procedia PDF Downloads 148

Authors: Eric Bang

Abstract:

Interactions between receptors and ligands are crucial for many essential biological processes, including neurotransmission and metabolism. Ligand-receptor analyses that examine cell behavior and interactions often utilize cell type-specific RNA expressions from single-cell RNA sequencing (scRNA-seq) data. Using CellPhoneDB, a public repository consisting of ligands, receptors, and ligand-receptor interactions, the cell-cell interactions were explored in a specific scRNA-seq dataset from kidney tissue and portrayed the results with dot plots and heat maps. Depending on the type of cell, each ligand-receptor pair was aligned with the interacting cell type and calculated the positori probabilities of these associations, with corresponding P values reflecting average expression values between the triads and their significance. Using single-cell data (sample kidney cell references), genes in the dataset were cross-referenced with ones in the existing CellPhoneDB dataset. For example, a gene such as Pleiotrophin (PTN) present in the single-cell data also needed to be present in the CellPhoneDB dataset. Using the single-cell transcriptomics data via slide-seq and reference data, the CellPhoneDB program defines cell types and plots them in different formats, with the two main ones being dot plots and heat map plots. The dot plot displays derived measures of the cell to cell interaction scores and p values. For the dot plot, each row shows a ligand-receptor pair, and each column shows the two interacting cell types. CellPhoneDB defines interactions and interaction levels from the gene expression level, so since the p-value is on a -log10 scale, the larger dots represent more significant interactions. By performing an interaction analysis, a significant interaction was discovered for myeloid and T-cell ligand-receptor pairs, including those between Secreted Phosphoprotein 1 (SPP1) and Fibronectin 1 (FN1), which is consistent with previous findings. It was proposed that an effective protocol would involve a filtration step where cell types would be filtered out, depending on which ligand-receptor pair is activated in that part of the tissue, as well as the incorporation of the CellPhoneDB data in a streamlined workflow pipeline. The filtration step would be in the form of a Python script that expedites the manual process necessary for dataset filtration. Being in Python allows it to be integrated with the CellPhoneDB dataset for future workflow analysis. The manual process involves filtering cell types based on what ligand/receptor pair is activated in kidney cells. One limitation of this would be the fact that some pairings are activated in multiple cells at a time, so the manual manipulation of the data is reflected prior to analysis. Using the filtration script, accurate sorting is incorporated into the CellPhoneDB database rather than waiting until the output is produced and then subsequently applying spatial data. It was envisioned that this would reveal wherein the cell various ligands and receptors are interacting with different cell types, allowing for easier identification of which cells are being impacted and why, for the purpose of disease treatment. The hope is this new computational method utilizing spatially explicit ligand-receptor association data can be used to uncover previously unknown specific interactions within kidney tissue.

Keywords: bioinformatics, Ligands, kidney tissue, receptors, spatial transcriptome

Procedia PDF Downloads 129

Authors: Stuart Noel

Abstract:

In a streetcar Named Desire, Tennessee Williams presents Blanche DuBois, a most complex and intriguing character who often uses prestige language to project the image of an upper-class speaker and to disguise her darker and complicated self. She embodies various fascinating and contrasting characteristics. Like New Orleans (the locale of the play), Blanche represents two opposing images. One image projects that of genteel, Southern charm and beauty, speaking formally and using prestige language and what some linguists refer to as “hypercorrection,” and the other image reveals that of a soiled, deteriorating façade, full of decadence and illusion. Williams said on more than one occasion that Blanche’s use of such language was a direct reflection of her personality and character (as a high school English teacher). Prestige language is an exaggeratedly elevated, pretentious, and oftentimes melodramatic form of one’s language incorporating superstandard or more standard speech than usual in order to project a highly authoritative individual identity. Speech styles carry personal identification meaning not only because they are closely associated with certain social classes but because they tend to be associated with certain conversational contexts. Features which may be considered to be “elaborated” in form (for example, full forms vs. contractions) tend to cluster together in speech registers/styles which are typically considered to be more formal and/or of higher social prestige, such as academic lectures and news broadcasts. Members of higher social classes have access to the elaborated registers which characterize formal writings and pre-planned speech events, such as lectures, while members of lower classes are relegated to using the more economical registers associated with casual, face-to-face conversational interaction, since they do not participate in as many planned speech events as upper-class speakers. Tennessee Williams’s work is characteristically concerned with the conflict between the illusions of an individual and the reality of his/her situation equated with a conflict between truth and beauty. An examination of Blanche DuBois reveals a recurring theme of art and decay and the use of prestige language to reveal artistry in language and to hide a deteriorating self. His graceful and poetic writing personifies her downfall and deterioration. Her loneliness and disappointment are the things so often strongly feared by the sensitive artists and heroes in the world. Hers is also a special and delicate human spirit that is often misunderstood and repressed by society. Blanche is afflicted with a psychic illness growing out of her inability to face the harshness of human existence. She is a sensitive, artistic, and beauty-haunted creature who is avoiding her own humanity while hiding behind her use of prestige language. And she embodies a partial projection of Williams himself.

Keywords: American drama, prestige language, Southern American literature, Tennessee Williams

Procedia PDF Downloads 360

Authors: Pengfei Meng, Shuangcheng Jia, Qian Li

Abstract:

We proposed a real-time high-precision semantic segmentation network based on a multi-resolution feature fusion module, the auxiliary feature extracting module, upsampling module, and atrous spatial pyramid pooling (ASPP) module. We designed a feature fusion structure, which is integrated with sufficient features of different resolutions. We also studied the effect of side-branch structure on the network and made discoveries. Based on the discoveries about the side-branch of the network structure, we used a side-branch auxiliary feature extraction layer in the network to improve the effectiveness of the network. We also designed upsampling module, which has better results than the original upsampling module. In addition, we also re-considered the locations and number of atrous spatial pyramid pooling (ASPP) modules and modified the network structure according to the experimental results to further improve the effectiveness of the network. The network presented in this paper takes the backbone network of Bisenetv2 as a basic network, based on which we constructed a network structure on which we made improvements. We named this network deep multiple-resolution bilateral networks for real-time, referred to as DMBR-Net. After experimental testing, our proposed DMBR-Net network achieved 81.2% mIoU at 119FPS on the Cityscapes validation dataset, 80.7% mIoU at 109FPS on the CamVid test dataset, 29.9% mIoU at 78FPS on the COCOStuff test dataset. Compared with all lightweight real-time semantic segmentation networks, our network achieves the highest accuracy at an appropriate speed.

Keywords: multi-resolution feature fusion, atrous convolutional, bilateral networks, pyramid pooling

Procedia PDF Downloads 134

Authors: Raed Alnajjar, Mohd Zakree, Ahmad Nazri

Abstract:

In this study, we apply Discrete Group Search Optimizer (DGSO) for solving Traveling Salesman Problem (TSP). The DGSO is a nature inspired optimization algorithm that imitates the animal behavior, especially animal searching behavior. The proposed DGSO uses a vector representation and some discrete operators, such as destruction, construction, differential evolution, swap and insert. The TSP is a well-known hard combinatorial optimization problem, which seeks to find the shortest path among numbers of cities. The performance of the proposed DGSO is evaluated and tested on benchmark instances which listed in LIBTSP dataset. The experimental results show that the performance of the proposed DGSO is comparable with the other methods in the state of the art for some instances. The results show that DGSO outperform Ant Colony System (ACS) in some instances whilst outperform other metaheuristic in most instances. In addition to that, the new results obtained a number of optimal solutions and some best known results. DGSO was able to obtain feasible and good quality solution across all dataset.

Keywords: discrete group search optimizer (DGSO); Travelling salesman problem (TSP); Variable neighborhood search(VNS)

Procedia PDF Downloads 313

Authors: Rosa Tsegaye Aga, Xuan Jiang, Pavel Vazquez Faci, Siqing Liu, Simon Rayner, Endalkachew Alemu, Markos Abebe

Abstract:

Tuberculosis (TB) is a major cause of disease globally. In most cases, TB is treatable and curable, but only with the proper treatment. There is a time when drug-resistant TB occurs when bacteria become resistant to the drugs that are used to treat TB. Current strategies to identify drug-resistant TB bacteria are laboratory-based, and it takes a longer time to identify the drug-resistant bacteria and treat the patient accordingly. But machine learning (ML) and data science approaches can offer new approaches to the problem. In this study, we propose to develop an ML-based model to predict the antibiotic resistance phenotypes of TB isolates in minutes and give the right treatment to the patient immediately. The study has been using the whole genome sequence (WGS) of TB isolates as training data that have been extracted from the NCBI repository and contain different countries’ samples to build the ML models. The reason that different countries’ samples have been included is to generalize the large group of TB isolates from different regions in the world. This supports the model to train different behaviors of the TB bacteria and makes the model robust. The model training has been considering three pieces of information that have been extracted from the WGS data to train the model. These are all variants that have been found within the candidate genes (F1), predetermined resistance-associated variants (F2), and only resistance-associated gene information for the particular drug. Two major datasets have been constructed using these three information. F1 and F2 information have been considered as two independent datasets, and the third information is used as a class to label the two datasets. Five machine learning algorithms have been considered to train the model. These are Support Vector Machine (SVM), Random forest (RF), Logistic regression (LR), Gradient Boosting, and Ada boost algorithms. The models have been trained on the datasets F1, F2, and F1F2 that is the F1 and the F2 dataset merged. Additionally, an ensemble approach has been used to train the model. The ensemble approach has been considered to run F1 and F2 datasets on gradient boosting algorithm and use the output as one dataset that is called F1F2 ensemble dataset and train a model using this dataset on the five algorithms. As the experiment shows, the ensemble approach model that has been trained on the Gradient Boosting algorithm outperformed the rest of the models. In conclusion, this study suggests the ensemble approach, that is, the RF + Gradient boosting model, to predict the antibiotic resistance phenotypes of TB isolates by outperforming the rest of the models.

Keywords: machine learning, MTB, WGS, drug resistant TB

Procedia PDF Downloads 35

Authors: Yuliya E. Leshchenko, Tatyana S. Ostapenko

Abstract:

Language variation is a widespread linguistic phenomenon which can affect different levels of a language system: phonological, morphological, lexical, syntactic, etc. It is obvious that the scope of possible standard alternations within a particular language is limited by a variety of its norms and regulations which set more or less clear boundaries for what is possible and what is not possible for the speakers. The possibility of lexical variation (alternate usage of lexical items within the same contexts) is based on the fact that the meanings of words are not clearly and rigidly defined in the consciousness of the speakers. Therefore, lexical variation is usually connected with unstable relationship between words and their referents: a case when a particular lexical item refers to different types of referents, or when a particular referent can be named by various lexical items. We assume that the scope of lexical variation in bilingual speech is generally wider than that observed in monolingual speech due to the fact that, besides ‘lexical item – referent’ relations it involves the possibility of cross-language variation of L1 and L2 lexical items. We use the term ‘cross-language variation’ to denote a case when two equivalent words of different languages are treated by a bilingual speaker as freely interchangeable within the common linguistic context. As distinct from code-switching which is traditionally defined as the conscious use of more than one language within one communicative act, in case of cross-language lexical variation the speaker does not perceive the alternate lexical items as belonging to different languages and, therefore, does not realize the change of language code. In the paper, the authors present research of lexical variation of adult Komi-Permyak – Russian bilingual speakers. The two languages co-exist on the territory of the Komi-Permyak District in Russia (Komi-Permyak as the ethnic language and Russian as the official state language), are usually acquired from birth in natural linguistic environment and, according to the data of sociolinguistic surveys, are both identified by the speakers as coordinate mother tongues. The experimental research demonstrated that alternation of Komi-Permyak and Russian words within one utterance/phrase is highly frequent both in speech perception and production. Moreover, our participants estimated cross-language word combinations like ‘маленькая /Russian/ нывка /Komi-Permyak/’ (‘a little girl’) or ‘мунны /Komi-Permyak/ домой /Russian/’ (‘go home’) as regular/habitual, containing no violation of any linguistic rules and being equally possible in speech as the equivalent intra-language word combinations (‘учöтик нывка’ /Komi-Permyak/ or ‘идти домой’ /Russian/). All the facts considered, we claim that constant concurrent use of the two languages results in the fact that a large number of their words tend to be intuitively interpreted by the speakers as lexical variants not only related to the same referent, but also referring to both languages or, more precisely, to none of them in particular. Consequently, we can suppose that bilingual mental lexicon includes an extensive ‘fused’ zone of lexical representations that provide the basis for cross-language variation in bilingual speech.

Keywords: bilingualism, bilingual mental lexicon, code-switching, lexical variation

Procedia PDF Downloads 139

Authors: Oladapo Tolulope Ibitoye

Abstract:

Neural networks have become prominent and widely engaged in algorithmic-based machine learning networks. They are perfect in solving day-to-day issues to a certain extent. Neural networks are computing systems with several interconnected nodes. One of the numerous areas of application of neural networks is object detection. This is a prominent area due to the coronavirus disease pandemic and the post-pandemic phases. Wearing a face mask in public slows the spread of the virus, according to experts’ submission. This calls for the development of a reliable and effective model for detecting face masks on people's faces during compliance checks. The existing neural network models for facemask detection are characterized by their black-box nature and large dataset requirement. The highlighted challenges have compromised the performance of the existing models. The proposed model utilized Faster R-CNN Model on Inception V3 backbone to reduce system complexity and dataset requirement. The model was trained and validated with very few datasets and evaluation results shows an overall accuracy of 96% regardless of skin tone.

Keywords: convolutional neural network, face detection, face mask, masked faces

Procedia PDF Downloads 54

Authors: Kenneth Fon Mbah, Arash Habibi Lashkari, Ali A. Ghorbani

Abstract:

Phishing e-mails are a security issue that not only annoys online users, but has also resulted in significant financial losses for businesses. Phishing advertisements and pornographic e-mails are difficult to detect as attackers have been becoming increasingly intelligent and professional. Attackers track users and adjust their attacks based on users’ attractions and hot topics that can be extracted from community news and journals. This research focuses on deceptive Phishing attacks and their variants such as attacks through advertisements and pornographic e-mails. We propose a framework called Phishing Alerting System (PHAS) to accurately classify e-mails as Phishing, advertisements or as pornographic. PHAS has the ability to detect and alert users for all types of deceptive e-mails to help users in decision making. A well-known email dataset has been used for these experiments and based on previously extracted features, 93.11% detection accuracy is obtainable by using J48 and KNN machine learning techniques. Our proposed framework achieved approximately the same accuracy as the benchmark while using this dataset.

Keywords: phishing e-mail, phishing detection, anti phishing, alarm system, machine learning

Procedia PDF Downloads 326