Search results for: CBCT images

Authors: Abe Degale D., Cheng Jian

Abstract:

When creating automated video surveillance systems, violent action recognition is crucial. In recent years, hand-crafted feature detectors have been the primary method for achieving violence detection, such as the recognition of fighting activity. Researchers have also looked into learning-based representational models. On benchmark datasets created especially for the detection of violent sequences in sports and movies, these methods produced good accuracy results. The Hockey dataset's videos with surveillance camera motion present challenges for these algorithms for learning discriminating features. Image recognition and human activity detection challenges have shown success with deep representation-based methods. For the purpose of detecting violent images and identifying aggressive human behaviours, this research suggested a deep representation-based model using the transfer learning idea. The results show that the suggested approach outperforms state-of-the-art accuracy levels by learning the most discriminating features, attaining 99.34% and 99.98% accuracy levels on the Hockey and Movies datasets, respectively.

Keywords: violence detection, faster RCNN, transfer learning and, surveillance video

Procedia PDF Downloads 106

Authors: Ali Berkan Ural

Abstract:

This paper presents the analysis, evaluation, and pre-diagnosis of early stage breast based diagnostic problems (breast cancer, nodulesorlumps) by Computer Aided Diagnosing (CAD) system from mammogram radiological images. According to the statistics, the time factor is crucial to discover the disease in the patient (especially in women) as possible as early and fast. In the study, a new algorithm is developed using advanced image processing and deep learning method to detect and classify the problem at earlystagewithmoreaccuracy. This system first works with image processing methods (Image acquisition, Noiseremoval, Region Growing Segmentation, Morphological Operations, Breast BorderExtraction, Advanced Segmentation, ObtainingRegion Of Interests (ROIs), etc.) and segments the area of interest of the breast and then analyzes these partly obtained area for cancer detection/lumps in order to diagnosis the disease. After segmentation, with using the Spectrogramimages, 5 different deep learning based methods (specified Convolutional Neural Network (CNN) basedAlexNet, ResNet50, VGG16, DenseNet, Xception) are applied to classify the breast based problems.

Keywords: computer aided diagnosis, breast cancer, region growing, segmentation, deep learning

Procedia PDF Downloads 95

Authors: Alisha Fisher

Abstract:

Sex trafficking is a worldwide crisis that requires trauma-informed and survivor-centered media attention to accurate disseminate information. Much of the previous literature on sex trafficking tends to focus on the frequency of incidents, intervention, and support strategies for survivors, with few of them looking to how the media is conducting their reporting on sex trafficking cases to the public. Utilizing data of reports from the media of cases of sex trafficking in the two Canadian provinces with the highest cases of sex trafficking, Ontario and Nova Scotia, the authors sought to analyze the similarities and differences of how sex trafficking cases were being reported. A total of twenty articles were examined, with ten based within the province of Ontario and the remaining ten from the province of Nova Scotia. The authors coded in two processes, first, who the article was about, and second, the framing and content inclusion. The results suggest that there is high usage and reliance of voices and images of authority, with male people of color being shown as the perpetrators and white women being shown as the survivors. These findings can aid in the expansion of trauma-informed, survivor-centered media literacy of reports of sex trafficking to provide accurate insights and further developing robust methods to intersectional approaches to reporting cases of sex trafficking.

Keywords: sex trafficking, media coverage, Canada sex trafficking, content analysis

Procedia PDF Downloads 189

Authors: Chanokporn Angsuviriya

Abstract:

The objective of the study is to analyse linguistic devices reflecting the violence in the south border provinces; namely Pattani, Yala, Narathiwat and Songkla on 1,344 front pages of three local newspapers; namely ChaoTai, Focus PhakTai and Samila Time and of two national newspapers, including ThaiRath and Matichon, between 2004 and 2005, and 2011 and 2012. The study shows that there are two important linguistic devices: 1) lexical choices consisting of the use of verbs describing violence, the use of quantitative words and the use of words naming someone who committed violent acts, and 2) metaphors consisting of “a violent problem is heat”, “a victim is a leaf”, and “a terrorist is a dog”. Comparing linguistic devices between two types of newspapers, national newspapers choose to use words more violently than local newspapers do. Moreover, they create more negative images of the south of Thailand by using stative verbs. In addition, in term of metaphors “a terrorist is a fox.” is only found in national newspapers. As regards naming terrorists “southern insurgents”, this noun phrase which is collectively called by national newspapers has strongly negative meaning. Moreover, “southern insurgents” have been perceived by the Thais in the whole country while “insurgents” that are not modified have been only used by local newspapers.

Keywords: linguistic devices, local newspapers, national newspapers, violence

Procedia PDF Downloads 241

Authors: Kamyar Fakhr, Roozbeh Salmani

Abstract:

Alphonse Bertillon developed the first biometric security system in the 1800s. Today, many governments and giant companies are considering or have procured biometrically enabled security schemes. Iris is a kaleidoscope of patterns and colors. Each individual holds a set of irises more unique than their thumbprint. Every single day, giant companies like Google and Apple are experimenting with reliable biometric systems. Now, after almost 200 years of improvements, face ID does not work with masks, it gives access to fake 3D images, and there is no global usage of biometric recognition systems as national identity (ID) card. The goal of this paper is to demonstrate the advantages of iris recognition overall biometric recognition systems. It make two extensions: first, we illustrate how a very large amount of internet fraud and cyber abuse is happening due to bugs in face recognition systems and in a very large dataset of 3.4M people; second, we discuss how establishing a secure global network of iris recognition devices connected to authoritative convolutional neural networks could be the safest solution to this dilemma. Another aim of this study is to provide a system that will prevent system infiltration caused by cyber-attacks and will block all wireframes to the data until the main user ceases the procedure.

Keywords: biometric system, convolutional neural network, cyber-attack, secure

Procedia PDF Downloads 218

Authors: Hemabushan K., Manikandan

Abstract:

In a swirl atomizer, fluid undergoes a swirling motion as a result of centrifugal force created by opposed tangential inlets in the swirl chamber. The angular momentum of fluid continually increases as it reaches the exit orifice and forms a hollow sheet. Which disintegrates to form ligaments and droplets respectively as it flows downstream. This type of atomizers used in rocket injectors and oil burner furnaces. In this present investigation a swirl atomizer with two opposed tangential inlets has been designed. Water as working fluid, experiments had been conducted for the fluid injection pressures in regime of 0.033 bar to 0.519 bar. The fluid has been pressured by a 0.5hp pump and regulated by a pressure regulator valve. Injection pressure of fluid has been measured by a U-tube mercury manometer. The spray pattern and the droplets has been captured with a high resolution camera in black background with a high intensity flash highlighting the fluid. The unprocessed images were processed in ImageJ processing software for measuring the droplet diameters and its shape characteristics along the downstream. The parameters such as mean droplet diameter and distribution, wave pattern, rupture distance and spray angle were studied for this atomizer. The above results were compared with theoretical results and also analysed for deviation with design parameters.

Keywords: swirl atomizer, injector, spray, SWIRL

Procedia PDF Downloads 490

Authors: K. M. Kheiralla, M. A. Ali, M. Y. Abdelgalil, N. E. Mohamed, G. Boutsis

Abstract:

The present study focus on integrated geophysical surveys carried out in the mineralization zone in Erkowit region, Eastern Sudan to determine the extensions of the potential ore deposits on the topographically high hilly area and under the cover of alluvium along the nearby wadi and to locate other occurrences if any. The magnetic method (MAG) and the electrical resistivity tomography (ERT) were employed for the survey. Eleven traverses were aligned approximately at right angles to the general strike of the rock formations. The disseminated sulfides are located on the alteration shear zone which is composed of granitic and dioritic highly ferruginated rock occupying the southwestern and central parts of the area, this was confirmed using thin and polished sections mineralogical analysis. The magnetic data indicates low magnetic values for wadi sedimentary deposits in its southern part of the area, and high anomalies which are suspected as gossans due to magnetite formed during wall rock alteration consequent to mineralization. The significant ERT images define low resistivity zone as traced as sheared zones which may associated with the main loci of ore deposition. The study designates that correlation of magnetic and ERT anomalies with lithology are extremely useful in mineral exploration due to variations in some specific physical properties of rocks.

Keywords: ERT, magnetic, mineralization, Red Sea, Sudan

Procedia PDF Downloads 389

Authors: Jorge R. Santos, Pedro Sebastiao

Abstract:

In recent years there has been a great growth in the use of drones, being used in several areas such as security, agriculture, or research. The existence of some systems that allow the remote control of drones is a reality; however, these systems are quite simple and directed to specific functionality. This paper proposes the development of a web platform made in Vue.js and Node.js to control, manage, and monitor drones in real time. Using a microservice architecture, the proposed project will be able to integrate algorithms that allow the optimization of processes. Communication with remote devices is suggested via HTTP through 3G, 4G, and 5G networks and can be done in real time or by scheduling routes. This paper addresses the case of forest fires as one of the services that could be included in a system similar to the one presented. The results obtained with the elaboration of this project were a success. The communication between the web platform and drones allowed its remote control and monitoring. The incorporation of the fire detection algorithm in the platform proved possible a real time analysis of the images captured by the drone without human intervention. The proposed system has proved to be an asset to the use of drones in fire detection. The architecture of the application developed allows other algorithms to be implemented, obtaining a more complex application with clear expansion.

Keywords: drone control, microservices, node.js, unmanned aerial vehicles, vue.js

Procedia PDF Downloads 148

Authors: Hadi Ardiny, Amir Mohammad Beigzadeh

Abstract:

Through the utilization of a combination of various sensors and data fusion methods, the detection of potential nuclear threats can be significantly enhanced by extracting more information from different data. In this research, an experimental and modeling approach was employed to track a radioactive source by combining a surveillance camera and a radiation detector (NaI). To run this experiment, three mobile robots were utilized, with one of them equipped with a radioactive source. An algorithm was developed in identifying the contaminated robot through correlation between camera images and camera data. The computer vision method extracts the movements of all robots in the XY plane coordinate system, and the detector system records the gamma-ray count. The position of the robots and the corresponding count of the moving source were modeled using the MCNPX simulation code while considering the experimental geometry. The results demonstrated a high level of accuracy in finding and locating the target in both the simulation model and experimental measurement. The modeling techniques prove to be valuable in designing different scenarios and intelligent systems before initiating any experiments.

Keywords: nuclear threats, radiation detector, MCNPX simulation, modeling techniques, intelligent systems

Procedia PDF Downloads 123

Authors: M. Dabirinezhad, M. Bayat Pour, A. Dabirinejad

Abstract:

In 2020, the technology was introduced to solve some of the deficiencies of direct digital radiology. SDDR is an invention that is capable of capturing dental images without human intervention, and it was invented by the authors of this paper. Adjusting the radiology wave dose is a part of the dentists, radiologists, and dental nurses’ tasks during the radiology photography process. In this paper, an improvement will be added to enable SDDR to set the suitable radiology wave dose according to the density and age of the patients automatically. The separate sensors will be included in the sensors’ package to use the ultrasonic wave to detect the density of the teeth and change the wave dose. It facilitates the process of dental photography in terms of time and enhances the accuracy of choosing the correct wave dose for each patient separately. Since the radiology waves are well known to trigger off other diseases such as cancer, choosing the most suitable wave dose can be helpful to decrease the side effect of that for human health. In other words, it decreases the exposure time for the patients. On the other hand, due to saving time, less energy will be consumed, and saving energy can be beneficial to decrease the environmental impact as well.

Keywords: dental direct digital imaging, environmental impacts, SDDR technology, wave dose

Procedia PDF Downloads 194

Authors: Soni Ariawan

Abstract:

The discourse of language and culture relation is an interesting issue to be researched. The debate is not about what comes first, language or culture, but it strongly argues that learning foreign language also means learning the culture of the language. The more interesting issue found once constructing an EFL textbook dealing with proportional representation among source culture, target culture and international culture. This study investigates cultural content representation in EFL textbook grade 10 year 2017 in Indonesia. Cortazzi and Jin’s theoretical framework is employed to analyse the reading texts, conversations, and images. The finding shows that national character as the main agenda of Indonesian government is revealed in this textbook since the textbook more frequently highlights the source culture (Indonesian culture) compared to target and international culture. This is aligned with the aim of Indonesian government to strengthen the national identity and promoting local culture awareness through education. To conclude, the study is expected to be significant in providing the idea for government to consider cultural balances representation in constructing textbook. Furthermore, teachers and students should be aware of cultural content revealed in the EFL textbook and be able to enhance intercultural communication not only in the classroom but also in a wider society.

Keywords: EFL textbook, intercultural communication, local culture, target culture, international culture

Procedia PDF Downloads 220

Authors: Mo-Se Lee, Cheol-Hwi Ahn, Kee-Young Kwahk, Hyunchul Ahn

Abstract:

Over the past decade, deep learning has been in spotlight among various machine learning algorithms. In particular, CNN (Convolutional Neural Network), which is known as effective solution for recognizing and classifying images, has been popularly applied to classification and prediction problems in various fields. In this study, we try to apply CNN to stock market prediction, one of the most challenging tasks in the machine learning research. In specific, we propose to apply CNN as the binary classifier that predicts stock market direction (up or down) by using a graph as its input. That is, our proposal is to build a machine learning algorithm that mimics a person who looks at the graph and predicts whether the trend will go up or down. Our proposed model consists of four steps. In the first step, it divides the dataset into 5 days, 10 days, 15 days, and 20 days. And then, it creates graphs for each interval in step 2. In the next step, CNN classifiers are trained using the graphs generated in the previous step. In step 4, it optimizes the hyper parameters of the trained model by using the validation dataset. To validate our model, we will apply it to the prediction of KOSPI200 for 1,986 days in eight years (from 2009 to 2016). The experimental dataset will include 14 technical indicators such as CCI, Momentum, ROC and daily closing price of KOSPI200 of Korean stock market.

Keywords: convolutional neural network, deep learning, Korean stock market, stock market prediction

Procedia PDF Downloads 425

Authors: Mohamed K. Metwally, Sherif H. El-Gohary, Kyung Min Byun, Seung Moo Han, Soo Yeol Lee, Min Hyoung Cho, Gon Khang, Jinsung Cho, Tae-Seong Kim

Abstract:

Photoacoustic imaging (PAI) is a non-invasive and non-ionizing imaging modality that combines the absorption contrast of light with ultrasound resolution. Laser is used to deposit optical energy into a target (i.e., optical fluence). Consequently, the target temperature rises, and then thermal expansion occurs that leads to generating a PA signal. In general, most image reconstruction algorithms for PAI assume uniform fluence within an imaging object. However, it is known that optical fluence distribution within the object is non-uniform. This could affect the reconstruction of PA images. In this study, we have investigated the influence of optical fluence distribution on PA back-propagation imaging using finite element method. The uniform fluence was simulated as a triangular waveform within the object of interest. The non-uniform fluence distribution was estimated by solving light propagation within a tissue model via Monte Carlo method. The results show that the PA signal in the case of non-uniform fluence is wider than the uniform case by 23%. The frequency spectrum of the PA signal due to the non-uniform fluence has missed some high frequency components in comparison to the uniform case. Consequently, the reconstructed image with the non-uniform fluence exhibits a strong smoothing effect.

Keywords: finite element method, fluence distribution, Monte Carlo method, photoacoustic imaging

Procedia PDF Downloads 377

Authors: Rasheed A. Saliu, Sunday E. Ogundipe, Oluwaseun A. Adefila

Abstract:

Information and Communication Technology (ICT) in recent time has transformed the means by which we inform ourselves, with world events and areas of personal interests, and further our learning. Today, for many, books and journals are no longer the first or primary source of information or learning. We now regularly rely on images, video, animations and sound to acquire information and to learn. Increased and improved access to the internet has accelerated this phenomenon. We now acquire and access information in ways fundamentally different from the pre-ICT era. But to what extent is academic staff in colleges of education, having access to and the utilising of ICT devices in their lecture deliveries especially in School of Science and Vocational and Technical? The main focus of this paper is to proffer solution to this salient question. It is essentially an empirical study carried out in five colleges of education in south-west zone of Nigeria. The target population was the academic staff in the selected institution. A total number of 150 male and female lecturers were contacted for the study. The main instrument was questionnaire. The finding reveals that male lecturers are much more ICT inclined than women folk in the academics. Some recommendations were made to endear academics to utilizing ICT at their disposal to foster qualitative delivery in this digital era.

Keywords: education, gender, ICT, Nigeria

Procedia PDF Downloads 297

Authors: Rachita Lal

Abstract:

Urban sprawl is a significant contributor to land use change in developing countries, where urbanization rates are high. The most important driver of environmental changes is also considered to be the shift in land use and land cover. Our local and regional land managers must carefully analyze urbanization and its effects on cities to make the best choices. This study uses satellite imagery to examine how urbanization affects the local ecosystem through geographic expansion. The following research focuses on the effects of city growth on the local environment, land use, and Land cover. The primary focus of this research is to study, To understand the role of urbanization on city expansion. To study the impact of spatial growth of urban areas on the Land cover. In this paper, the GIS tool will be used to analyze. For this purpose, four digital images are used for the years 2000, 2005, 2011, and 2019. The use of the approach in the Bhubaneswar Urban Core, one of the fastest developing and planned cities in India, has proved that it is highly beneficial and successful for monitoring urban sprawl. It offers a helpful tool for quantitative assessment, which is crucial for determining the spatial dynamics, variations, and changes of urban sprawl patterns in quickly increasing regions.

Keywords: LULC, urbanization, environment impact assessment, spatial growth

Procedia PDF Downloads 120

Authors: Kerrin Burnell

Abstract:

In a world of increasingly technology-dependent students, the English language classroom should ideally keep up with developments to keep students engaged as much as possible. Unfortunately, as is the case in Oman, funding is not always adequate to ensure students have the most up to date technology, and most institutions are still reliant on paper-based textbooks. In order to try and bridge the gap between the technology available (smartphones) and textbooks, augmented reality (AR) technology can be utilized to enhance classroom, homework, and extracurricular activities. AR involves overlaying media (videos, images etc) over the top of physical objects (posters, book pages etc) and then sharing the media. This case study involved introducing students to a freely available entry level AR app called Aurasma. Students were asked to augment their English textbooks, word walls, research project posters, and extracurricular posters. Through surveys, interviews and an analysis of time spent accessing the different media, a determination of the appropriateness of the technology for the classroom was determined. Results indicate that the use of AR has positive effects on many aspects of the English classroom. Increased student engagement, total time spent on task, interaction, and motivation were evident, along with a decrease in technology-related anxiety. As it is proving very difficult to get tablets or even laptops in classrooms in Oman, these preliminary results indicate that many positive outcomes will come from introducing students to this innovative technology.

Keywords: augmented reality, classroom technology, classroom innovation, engagement

Procedia PDF Downloads 382

Authors: Su-Ning Wang, Yao-Qiang Chen

Abstract:

The CeO2-ZrO2-La2O3-Al2O3 composite oxides are synthesized using co-precipitation method by two different reactors (i.e. continuous stirred-tank reactor and batch reactor), and the corresponding Rh-only three-way catalysts are obtained by wet-impregnation approach. The textural, structural, morphology and redox properties of the support materials, as well as the catalytic performance of the Rh-only catalyst are investigated systematically. The results reveal that the materials (CZLA-C) synthesized by continuous stirred-tank reactor have a better physic-chemical properties than the counterpart material (CZLA-B) prepared by batch reactor. After aging treatment at 1000 ℃ for 5 h, the BET surface area and pore volume of S1 reach up to 76 m2 g-1 and 0.36 mL/g, respectively, which is higher than that of S2. The XRD and Raman results demonstrate that a high structural stability is obtained by S1 because of the negligible lattice variation and the slight grain growth after aging treatment. The SEM and TEM images display that the morphology of S1 is assembled by many homogeneous primary nanoparticles (about 6.12 nm) that are connected to form mesoporous structure The TPR measurement shows that S1 possesses a higher reduction ability than S2. Compared with the catalyst supported on the CZLA-B, the as-prepared CZLA-C demonstrates an improved three-way catalytic activity both before and after aging treatment.

Keywords: composite oxides, reactor, catalysis, catalytic performance

Procedia PDF Downloads 297

Authors: E. Solfaroli Camillocci, C. Mancini-Terracciano, V. Bocci, A. Carollo, M. Colandrea, F. Collamati, M. Cremonesi, M. E. Ferrari, P. Ferroli, F. Ghielmetti, C. M. Grana, M. Marafini, S. Morganti, M. Patane, G. Pedroli, B. Pollo, L. Recchia, A. Russomando, M. Schiariti, M. Toppi, G. Traini, R. Faccini

Abstract:

A Radio-Guided Surgery technique exploiting β− emitting radio-tracers has been suggested to overcome the impact of the large penetration of γ radiation. The detection of electrons in low radiation background provides a clearer delineation of the margins of lesioned tissues. As a start, the clinical cases were selected between the tumors known to express receptors to a β− emitting radio-tracer: 90Y-labelled DOTATOC. The results of tests on ex-vivo specimens of meningioma brain tumor and abdominal neuroendocrine tumors are presented. Voluntary patients were enrolled according to the standard uptake value (SUV > 2 g/ml) and the expected tumor-to-non-tumor ratios (TNR∼10) estimated from PET images after administration of 68Ga-DOTATOC. All these tests validated this technique yielding a significant signal on the bulk tumor and a negligible background from the nearby healthy tissue. Even injecting as low as 1.4 MBq/kg of radiotracer, tumor remnants of 0.1 ml would be detectable. The negligible medical staff exposure was confirmed and among the biological wastes only urine had a significant activity.

Keywords: ex-vivo test, meningioma, neuroendocrine tumor, radio-guided surgery

Procedia PDF Downloads 294

Authors: Sanam Pudasaini, A. T. K. Perera, Ahmed Syed Shaheer Uddin, Sum Huan Ng, Chun Yang

Abstract:

Development of high-efficiency and environment friendly bacterial inactivation methods is of great importance for preventing waterborne diseases which are one of the leading causes of death in the world. Traditional bacterial inactivation methods (e.g., ultraviolet radiation and chlorination) have several limitations such as longer treatment time, formation of toxic byproducts, bacterial regrowth, etc. Recently, an electroporation-based inactivation method was introduced as a substitute. Here, an electroporation-based continuous flow microfluidic device equipped with an array of micropillars is developed, and the device achieved high bacterial inactivation performance ( > 99.9%) within a short exposure time ( < 1 s). More than 99.9% reduction of Escherichia coli bacteria was obtained for the flow rate of 1 mL/hr, and no regrowth of bacteria was observed. Images from scanning electron microscope confirmed the formation of electroporation-induced nano-pore within the cell membrane. Through numerical simulation, it has been shown that sufficiently large electric field strength (3 kV/cm), required for bacterial electroporation, were generated using PDMS micropillars for an applied voltage of 300 V. Further, in this method of inactivation, there is no involvement of chemicals and the formation of harmful by-products is also minimum.

Keywords: electroporation, high-efficiency, inactivation, microfluidics, micropillar

Procedia PDF Downloads 180

Authors: Alfred Chen, Cheng Chieh Hsu, Yu-Pin Ma, Meng-Jie Lin, Fu Pai Chiu, Yi-Yan Sie

Abstract:

This study is attempted to construct a human-computer interactive platform system that has mainly consisted of an augmented hardware system, a software system, a display table, and mixed media. This system has provided with human-computer interaction services through an interactive platform for the tourism industry. A well designed interactive platform, integrating of augmented reality and mixed media, has potential to enhance museum display quality and diversity. Besides, it will create a comprehensive and creative display mode for most museums and historical heritages. Therefore, it is essential to let public understand what the platform is, how it functions, and most importantly how one builds an interactive augmented platform. Hence the authors try to elaborate the construction process of the platform in detail. Thus, there are three issues to be considered, i.e.1) the theory and application of augmented reality, 2) the hardware and software applied, and 3) the mixed media presented. In order to describe how the platform works, Courtesy Door of Tainan Confucius Temple has been selected as case study in this study. As a result, a developed interactive platform has been presented by showing the physical entity object, along with virtual mixing media such as text, images, animation, and video. This platform will result in providing diversified and effective information that will be delivered to the users.

Keywords: human-computer interaction, mixed reality, mixed media, tourism

Procedia PDF Downloads 489

Authors: Kaushik Sathupadi, Sandesh Achar

Abstract:

Human action recognition modeling is a critical task in machine learning. These systems require better techniques for recognizing body parts and selecting optimal features based on vision sensors to identify complex action patterns efficiently. Still, there is a considerable gap and challenges between images and videos, such as brightness, motion variation, and random clutters. This paper proposes a robust approach for classifying human actions over cloud-based image data. First, we apply pre-processing and detection, human and outer shape detection techniques. Next, we extract valuable information in terms of cues. We extract two distinct features: fuzzy local binary patterns and sequence representation. Then, we applied a greedy, randomized adaptive search procedure for data optimization and dimension reduction, and for classification, we used a random forest. We tested our model on two benchmark datasets, AAMAZ and the KTH Multi-view football datasets. Our HMR framework significantly outperforms the other state-of-the-art approaches and achieves a better recognition rate of 91% and 89.6% over the AAMAZ and KTH multi-view football datasets, respectively.

Keywords: computer vision, human motion analysis, random forest, machine learning

Procedia PDF Downloads 36

Authors: Musatafa Abbas Abbood Albadr, Masri Ayob, Sabrina Tiun, Fahad Taha Al-Dhief, Mohammad Kamrul Hasan

Abstract:

Breast Cancer (BC) is considered one of the most frequent reasons of cancer death in women between 40 to 55 ages. The BC is diagnosed by using digital images of the FNA (Fine Needle Aspirate) for both benign and malignant tumors of the breast mass. Therefore, this work proposes the Online Sequential Extreme Learning Machine (OSELM) algorithm for diagnosing BC by using the tumor features of the breast mass. The current work has used the Wisconsin Diagnosis Breast Cancer (WDBC) dataset, which contains 569 samples (i.e., 357 samples for benign class and 212 samples for malignant class). Further, numerous measurements of assessment were used in order to evaluate the proposed OSELM algorithm, such as specificity, precision, F-measure, accuracy, G-mean, MCC, and recall. According to the outcomes of the experiment, the highest performance of the proposed OSELM was accomplished with 97.66% accuracy, 98.39% recall, 95.31% precision, 97.25% specificity, 96.83% F-measure, 95.00% MCC, and 96.84% G-Mean. The proposed OSELM algorithm demonstrates promising results in diagnosing BC. Besides, the performance of the proposed OSELM algorithm was superior to all its comparatives with respect to the rate of classification.

Keywords: breast cancer, machine learning, online sequential extreme learning machine, artificial intelligence

Procedia PDF Downloads 111

Authors: Felix Boehnisch, Alexander Lutz

Abstract:

Based on the number of users and the amount of content posted daily, Facebook is considered the largest social network in the world. This content includes images or text posts from companies but also private persons, which are also commented on by other users. However, it can sometimes be difficult for companies to keep track of all the posts and the reactions to them, especially when there are several posts a day that contain hundreds to thousands of comments. To facilitate this, the following paper deals with the possible applications of sentiment analysis to social media comments in order to be able to support the work in social media marketing. In a first step, post comments were divided into positive and negative by a subjective rating, then the same comments were checked for their polarity value by the two german python libraries TextBlobDE and SentiWS and also grouped into positive, negative, or even neutral. As a control, the subjective classifications were compared with the machine-generated ones by a confusion matrix, and relevant quality criteria were determined. The accuracy of both libraries was not really meaningful, with 60% to 66%. However, many words or sentences were not evaluated at all, so there seems to be room for optimization to possibly get more accurate results. In future studies, the use of these specific German libraries can be optimized to gain better insights by either applying them to stricter cleaned data or by adding a sentiment value to emojis, which have been removed from the comments in advance, as they are not contained in the libraries.

Keywords: Facebook, German libraries, polarity, sentiment analysis, social media comments

Procedia PDF Downloads 182

Authors: Bandar Alahmadi, Lethia Jackson

Abstract:

Machine Learning model is used today in many real-life applications. The safety and security of such model is important, so the results of the model are as accurate as possible. One challenge of machine learning model security is the adversarial examples attack. Adversarial examples are designed by the attacker to cause the machine learning model to misclassify the input. We propose a method to generate adversarial examples to attack image classifiers. We are modifying the successfully classified images, so a classifier misclassifies them after the modification. In our method, we do not update the whole image, but instead we detect the important region, modify it, place it back to the original image, and then run it through a classifier. The algorithm modifies the detected region using two methods. First, it will add abstract image matrix on back of the detected image matrix. Then, it will perform a rotation attack to rotate the detected region around its axes, and embed the trace of image in image background. Finally, the attacked region is placed in its original position, from where it was removed, and a smoothing filter is applied to smooth the background with foreground. We test our method in cascade classifier, and the algorithm is efficient, the classifier confident has dropped to almost zero. We also try it in CNN (Convolutional neural network) with higher setting and the algorithm was successfully worked.

Keywords: adversarial examples, attack, computer vision, image processing

Procedia PDF Downloads 339

Authors: Yi-Feng Kao, Huey-Jine Chai, Chin-I Chang, Yi-Chen Chen, June-Ru Chen

Abstract:

Microglia is a macrophage that resides in brain, and overactive microglia may result in brain neuron damage or inflammation. In this study, the phospholipids was extracted from squid skin and manufactured into a liposome (SQ liposome) to mimic apoptotic body. We then evaluated anti-inflammatory effects of SQ liposome on mouse microglial cell line (BV-2) by lipopolysaccharide (LPS) induction. First, the major phospholipid constituents in the squid skin extract were including 46.2% of phosphatidylcholine, 18.4% of phosphatidylethanolamine, 7.7% of phosphatidylserine, 3.5% of phosphatidylinositol, 4.9% of Lysophosphatidylcholine and 19.3% of other phospholipids by HPLC-UV analysis. The contents of eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) in the squid skin extract were 11.8 and 28.7%, respectively. The microscopic images showed that microglia cells can engulf apoptotic cells or SQ-liposome. In cell based studies, there was no cytotoxicity to BV-2 as the concentration of SQ-liposome was less than 2.5 mg/mL. The LPS induced pro-inflammatory cytokines, including tumor necrosis factor-alpha (TNF-α) and interleukin-6 (IL-6), were significant suppressed (P < 0.05) by pretreated 0.03~2.5mg/ml SQ liposome. Oppositely, the anti-inflammatory cytokines transforming growth factor-beta (TGF-β) and interleukin-10 (IL-10) secretion were enhanced (P < 0.05). The results suggested that SQ-liposome possess anti-inflammatory properties on BV-2 and may be a good strategy for against neuro-inflammatory disease.

Keywords: apoptotic mimicry, neuroinflammation, microglia, squid processing by-products

Procedia PDF Downloads 482

Authors: M.F. Elkady, Sahar Zaki, Desouky Abd-El-Haleem

Abstract:

Silver nanoparticles (AgNPs) are already widely prepared using different technologies. However, there are limited data on the effects of hydrogen ion concentration on nano-silver production. In this investigation, the impact of the pH reaction medium toward the particle size, agglomeration and the yield of the produced bio-synthesized silver were established. Quasi-spherical silver nanoparticles were synthesized through the biosynthesis green production process using the Egyptian E. coli bacterial strain 23N at different pH values. The formation of AgNPs has been confirmed with ultraviolet–visible spectra through identification of their characteristic peak at 410 nm. The quantitative production yield and the orientation planes of the produced nano-silver were examined using X-ray spectroscopy (EDS) and X-ray diffraction (XRD). Quantitative analyses indicated that the silver production yield was promoted at elevated pH regarded to increase the reduction rate of silver precursor through both chemical and biological processes. As a result, number of the nucleus and thus the size of the silver nanoparticles were tunable through changing pH of the reaction system. Accordingly, the morphological structure and size of the produced silver and its aggregates were determined using scanning electron microscopy (SEM) and transmission electron microscopy (TEM) images. It was considered that the increment in pH value of the reaction media progress the aggregation of silver clusters. However, the presence of stain 23N biomass decreases the possibility of silver aggregation at the pH 7.

Keywords: silver nanoparticles, biosynthesis, reaction media pH, nano-silver characterization

Procedia PDF Downloads 371

Authors: Miho Tsukamoto

Abstract:

Since the initial creation of the Barbie doll in 1959, it became a symbol of US society. Likewise, the Licca-chan, a Japanese doll created in 1967, also became a Japanese symbolic doll of Japanese society. Prior to the introduction of Licca-chan, Barbie was already marketed in Japan but their sales were dismal. Licca-chan (an actual name: Kayama Licca) is a plastic doll with a variety of sizes ranging from 21.0 cm to 29.0 cm which many Japanese girls dream of having. For over 35 years, the manufacturer, Takara Co., Ltd. has sold over 48 million dolls and has produced doll houses, accessories, clothes, and Licca-chan video games for the Nintendo DS. Many First-generation Licca-chan consumers still are enamored with Licca-chan, and go to Licca-chan House, in an amusement park with their daughters. These people are called Licca-chan maniacs, as they enjoy touring the Licca-chan’s factory in Tohoku or purchase various Licca-chan accessories. After the successful launch of Licca-chan into the Japanese market, a mixed-like doll from the US and Japan, a doll, JeNny, was later sold in the same Japanese market by Takara Co., Ltd. in 1982. Comparison of these cultural iconic dolls, Barbie and Licca-chan, are analyzed in this paper. In fact, these dolls have concepts of girls’ dreams. By using concepts of mythology of Jean Baudrillard, these dolls can be represented idealized images of figures in the products for consumers, but at the same time, consumers can see products with different perspectives, which can cause controversy.

Keywords: Barbie, dolls, JeNny, idealization, Licca-chan

Procedia PDF Downloads 271

Authors: Hao Yuan, Yongjian Shen, Xiangjun He, Yuheng Li, Zhouzhou Zhang, Pengyu Zhang, Minkang Cai

Abstract:

Remote sensing data can provide support for decision-making in disaster assessment or disaster relief. The traditional processing methods of sensitive targets in remote sensing mapping are mainly based on manual retrieval and image editing tools, which are inefficient. Methods based on deep learning for sensitive target hiding are faster and more flexible. But these methods have disadvantages in training time and cost of calculation. This paper proposed a target hiding model Self Attention (SA) Deepfill, which used self-attention modules to replace part of gated convolution layers in image inpainting. By this operation, the calculation amount of the model becomes smaller, and the performance is improved. And this paper adds free-form masks to the model’s training to enhance the model’s universal. The experiment on an open remote sensing dataset proved the efficiency of our method. Moreover, through experimental comparison, the proposed method can train for a longer time without over-fitting. Finally, compared with the existing methods, the proposed model has lower computational weight and better performance.

Keywords: remote sensing mapping, image inpainting, self-attention mechanism, target hiding

Procedia PDF Downloads 136

Authors: Nikolaos Papadakis, Constantinos Condaxakis, Konstantinos Savvakis

Abstract:

The aim of this paper is to determine the elastic properties (elastic modulus and Poisson ratio) of a composite material based on noncontacting imaging methods. More specifically, the significantly reduced cost of digital cameras has given the opportunity of the high reliability of low-cost strain measurement. The open source platform Ncorr is used in this paper which utilizes the method of digital image correlation (DIC). The use of digital image correlation in measuring strain uses random speckle preparation on the surface of the gauge area, image acquisition, and postprocessing the image correlation to obtain displacement and strain field on surface under study. This study discusses technical issues relating to the quality of results to be obtained are discussed. [0]8 fabric glass/epoxy composites specimens were prepared and tested at different orientations 0[o], 30[o], 45[o], 60[o], 90[o]. Each test was recorded with the camera at a constant frame rate and constant lighting conditions. The recorded images were processed through the use of the image processing software. The parameters of the test are reported. The strain map output which is obtained through strain measurement using Ncorr is validated by a) comparing the elastic properties with expected values from Classical laminate theory, b) through finite element analysis.

Keywords: composites, Ncorr, strain map, videoextensometry

Procedia PDF Downloads 144

Authors: Asiri Wijesinghe, N. D. Kodikara, Damitha Sandaruwan

Abstract:

The Diabetic Retinopathy (DR) is a rapidly growing interrogation around the world which can be annotated by abortive metabolism of glucose that causes long-term infection in human retina. This is one of the preliminary reason of visual impairment and blindness of adults. Information on retinal pathological mutation can be recognized using ocular fundus images. In this research, we are mainly focused on resurrecting an automated diagnosis system to detect DR anomalies such as severity level classification of DR patient (Non-proliferative Diabetic Retinopathy approach) and vessel tortuosity measurement of untwisted vessels to assessment of vessel anomalies (Proliferative Diabetic Retinopathy approach). Severity classification method is obtained better results according to the precision, recall, F-measure and accuracy (exceeds 94%) in all formats of cross validation. In ROC (Receiver Operating Characteristic) curves also visualized the higher AUC (Area Under Curve) percentage (exceeds 95%). User level evaluation of severity capturing is obtained higher accuracy (85%) result and fairly better values for each evaluation measurements. Untwisted vessel detection for tortuosity measurement also carried out the good results with respect to the sensitivity (85%), specificity (89%) and accuracy (87%).

Keywords: fundus image, exudates, microaneurisms, hemorrhages, tortuosity, diabetic retinopathy, optic disc, fovea

Procedia PDF Downloads 341