Search results for: convolutional long short-term memory

Authors: Harsh Sadawarti, Kamal Malik

Abstract:

Humans are the unique and the most powerful creature on this planet just because of the high level of intelligence gifted by nature. Computational Intelligence is highly influenced by the term natural intelligence, neurosciences and mathematics. To deal with the in-depth study of computational intelligence and to utilize it in real-life applications, it is quite important to understand its simulation with the human brain. In this paper, the three important parts, Frontal Lobe, Occipital Lobe and Parietal Lobe of the human brain, are compared with the ANN(Artificial Neural Network), CNN(Convolutional Neural network), and RNN(Recurrent Neural Network), respectively. Intelligent computational systems are created by combining deductive reasoning, logical concepts and high-level algorithms with the simulation and study of the human brain. Human brain is a combination of Physiology, Psychology, emotions, calculations and many other parameters which are of utmost importance that determines the overall intelligence. To create intelligent algorithms, smart machines and to simulate the human brain in an effective manner, it is quite important to have an insight into the human brain and the basic concepts of biological neurosciences.

Keywords: computational intelligence, neurosciences, convolutional neural network, recurrent neural network, artificial neural network, frontal lobe, occipital lobe, parietal lobe

Procedia PDF Downloads 111

Authors: Despoina Mouratidis, Katia Lida Kermanidis

Abstract:

Machine translation (MT) is a very important task in Natural Language Processing (NLP). MT evaluation is crucial in MT development, as it constitutes the means to assess the success of an MT system, and also helps improve its performance. Several methods have been proposed for the evaluation of (MT) systems. Some of the most popular ones in automatic MT evaluation are score-based, such as the BLEU score, and others are based on lexical similarity or syntactic similarity between the MT outputs and the reference involving higher-level information like part of speech tagging (POS). This paper presents a language-independent machine learning framework for classifying pairwise translations. This framework uses vector representations of two machine-produced translations, one from a statistical machine translation model (SMT) and one from a neural machine translation model (NMT). The vector representations consist of automatically extracted word embeddings and string-like language-independent features. These vector representations used as an input to a multi-layer neural network (NN) that models the similarity between each MT output and the reference, as well as between the two MT outputs. To evaluate the proposed approach, a professional translation and a "ground-truth" annotation are used. The parallel corpora used are English-Greek (EN-GR) and English-Italian (EN-IT), in the educational domain and of informal genres (video lecture subtitles, course forum text, etc.) that are difficult to be reliably translated. They have tested three basic deep learning (DL) architectures to this schema: (i) fully-connected dense, (ii) Convolutional Neural Network (CNN), and (iii) Long Short-Term Memory (LSTM). Experiments show that all tested architectures achieved better results when compared against those of some of the well-known basic approaches, such as Random Forest (RF) and Support Vector Machine (SVM). Better accuracy results are obtained when LSTM layers are used in our schema. In terms of a balance between the results, better accuracy results are obtained when dense layers are used. The reason for this is that the model correctly classifies more sentences of the minority class (SMT). For a more integrated analysis of the accuracy results, a qualitative linguistic analysis is carried out. In this context, problems have been identified about some figures of speech, as the metaphors, or about certain linguistic phenomena, such as per etymology: paronyms. It is quite interesting to find out why all the classifiers led to worse accuracy results in Italian as compared to Greek, taking into account that the linguistic features employed are language independent.

Keywords: machine learning, machine translation evaluation, neural network architecture, pairwise classification

Procedia PDF Downloads 132

Authors: Gulcan Ozerim, Gunay Anlas

Abstract:

In shape memory alloys, upon loading, stress increases around crack tip and a martensitic phase transformation occurs in early stages. In many studies the stress distribution in the vicinity of the crack tip is represented by using linear elastic fracture mechanics (LEFM) although the pseudo-elastic behavior results in a nonlinear stress-strain relation. In this study, the HRR singularity (Hutchinson, Rice and Rosengren), that uses Rice’s path independent J-integral, is tried to formulate the stress distribution around the crack tip. In HRR approach, the Ramberg-Osgood model for the stress-strain relation of power-law hardening materials is used to represent the elastic-plastic behavior. Although it is recoverable, the inelastic portion of the deformation in martensitic transformation (up to the end of transformation) resembles to that of plastic deformation. To determine the constants of the Ramberg-Osgood equation, the material’s response is simulated in ABAQUS using a UMAT based on ZM (Zaki-Moumni) thermo-mechanically coupled model, and the stress-strain curve of the material is plotted. An edge cracked shape memory alloy (Nitinol) plate is loaded quasi-statically under mode I and modeled using ABAQUS; the opening stress values ahead of the cracked tip are calculated. The stresses are also evaluated using the asymptotic equations of both LEFM and HRR. The results show that in the transformation zone around the crack tip, the stress values are much better represented when the HRR singularity is used although the J-integral does not show path independent behavior. For the nodes very close to the crack tip, the HRR singularity is not valid due to the non-proportional loading effect and high-stress values that go beyond the transformation finish stress.

Keywords: crack, HRR singularity, shape memory alloys, stress distribution

Procedia PDF Downloads 325

Authors: M. M. Essa, S. Subash, M. Akbar, S. Al-Adawi, A. Al-Asmi, G. J. Guillemein

Abstract:

Objective: Transgenic (tg) mice which contain an amyloid precursor protein (APP) gene mutation, develop extracellular amyloid beta (Aβ) deposition in the brain, and severe memory and behavioural deficits with age. These mice serve as an important animal model for testing the efficacy of novel drug candidates for the treatment and management of symptoms of Alzheimer's disease (AD). Several reports have suggested that oxidative stress is the underlying cause of Aβ neurotoxicity in AD. Pomegranates contain very high levels of antioxidants and several medicinal properties that may be useful for improving the quality of life in AD patients. In this study, we investigated the effect of dietary supplementation of Omani pomegranate extract on the memory, anxiety and learning skills along with inflammation in an AD mouse model containing the double Swedish APP mutation (APPsw/Tg2576). Methods: The experimental groups of APP-transgenic mice from the age of 4 months were fed custom-mix diets (pellets) containing 4% pomegranate. We assessed spatial memory and learning ability, psychomotor coordination, and anxiety-related behavior in Tg and wild-type mice at the age of 4-5 months and 18-19 months using the Morris water maze test, rota rod test, elevated plus maze test, and open field test. Further, inflammatory parameters also analysed. Results: APPsw/Tg2576 mice that were fed a standard chow diet without pomegranates showed significant memory deficits, increased anxiety-related behavior, and severe impairment in spatial learning ability, position discrimination learning ability and motor coordination along with increased inflammation compared to the wild type mice on the same diet, at the age of 18-19 months In contrast, APPsw/Tg2576 mice that were fed a diet containing 4% pomegranates showed a significant improvements in memory, learning, locomotor function, and anxiety with reduced inflammatory markers compared to APPsw/Tg2576 mice fed the standard chow diet. Conclusion: Our results suggest that dietary supplementation with pomegranates may slow the progression of cognitive and behavioural impairments in AD. The exact mechanism is still unclear and further extensive research needed.

Keywords: Alzheimer's disease, pomegranates, oman, cognitive decline, memory loss, anxiety, inflammation

Procedia PDF Downloads 528

Authors: Muhammad Atif, Cang Yan

Abstract:

The enhancement of low-light images is a significant area of study aimed at enhancing the quality of captured images in challenging lighting environments. Recently, methods based on convolutional neural networks (CNN) have gained prominence as they offer state-of-the-art performance. However, many approaches based on CNN rely on increasing the size and complexity of the neural network. In this study, we propose an alternative method for improving low-light images using an autoencoder-based multiscale knowledge transfer model. Our method leverages the power of three autoencoders, where the encoders of the first two autoencoders are directly connected to the decoder of the third autoencoder. Additionally, the decoder of the first two autoencoders is connected to the encoder of the third autoencoder. This architecture enables effective knowledge transfer, allowing the third autoencoder to learn and benefit from the enhanced knowledge extracted by the first two autoencoders. We further integrate the proposed model into the PIX to PIX GAN framework. By integrating our proposed model as the generator in the GAN framework, we aim to produce enhanced images that not only exhibit improved visual quality but also possess a more authentic and realistic appearance. These experimental results, both qualitative and quantitative, show that our method is better than the state-of-the-art methodologies.

Keywords: low light image enhancement, deep learning, convolutional neural network, image processing

Procedia PDF Downloads 80

Authors: Ahmed Emara, Maged Ghoneima, Mohamed Dessouky

Abstract:

Demand for low power fast memories is increasing with the increase in IC’s complexity, in this paper we introduce a proposal for a compact SRAM based on memristor devices. The compact size of the proposed cell (1T2M compared to 6T of traditional SRAMs) allows denser memories on the same area. In this paper, we will discuss the proposed memristor memory cell for single/multi bit data storing configurations along with the writing and reading operations. Stored data stability across successive read operation will be illustrated, operational simulation results and a comparison of our proposed design with previously conventional SRAM and previously proposed memristor cells will be provided.

Keywords: memristor, multi-bit, single-bit, circuits, systems

Procedia PDF Downloads 374

Authors: Cory B. Lord

Abstract:

Falls among the elderly population has become an area of concern in healthcare today. The negative impacts of falls lead to increased morbidity, mortality, and financial burdens for both patients and healthcare systems. Falls in the United States is reported at an annual rate of 36 million in those aged 65 and older. Each year, one out of four people in this age group will suffer a fall, with 20% of these falls causing injury. The setting for this Doctor of Nursing Practice (DNP) project was a memory care unit in an assisted living community, as these facilities house cognitively impaired older adults. These communities lack fall prevention programs; therefore, the need exists to add to the body of knowledge to positively impact this population. The objective of this project was to reduce fall rates through the implementation of the Center for Disease Control and Prevention (CDC) STEADI (stopping elderly accidents, deaths, and injuries) program. The DNP project performed was a quality improvement pilot study with a pre and post-test design. This program was implemented in the memory care setting over 12 weeks. The project included an educational session for staff and a fall risk assessment with appropriate resident referrals. The three aims of the DNP project were to reduce fall rates among the elderly aged 65 and older who reside in the memory care unit, increase staff knowledge of STEADI fall prevention measures after an educational session, and assess the willingness of memory care unit staff to adopt an evidence-based a fall prevention program. The Donabedian model was used as a guiding conceptual framework for this quality improvement pilot study. The fall rate data for 12 months before the intervention was evaluated and compared to post-intervention fall rates. The educational session comprised of a pre and post-test to assess staff knowledge of the fall prevention program and the willingness of staff to adopt the fall prevention program. The overarching goal was to reduce falls in the elderly population who live in memory care units. The results of the study showed, on average that the fall rate during the implementation period of STEADI (μ=6.79) was significantly lower when compared to the prior 12 months (μ= 9.50) (p=0.02, α = 0.05). The mean staff knowledge scores improved from pretest (μ=77.74%) to post-test (μ=87.42%) (p=0.00, α= 0.05) after the education session. The results of the willingness to adopt a fall prevention program were scored at 100%. In summation, implementing the STEADI fall prevention program can assist in reducing fall rates for residents aged 65 and older who reside in a memory care setting.

Keywords: dementia, elderly, falls, STEADI

Procedia PDF Downloads 129

Authors: Aikaterini Kentroti, Theodore H. Kaskalis

Abstract:

Today, web servers are expected to serve thousands of client requests concurrently within stringent response time limits. In this paper, we evaluate experimentally and compare the performance as well as the resource utilization of popular web servers, which differ in their approach to handle concurrency. More specifically, Central Processing Unit (CPU)- and I/O intensive tests were conducted against the thread-based Apache and Go as well as the event-based Nginx and Node.js under increasing concurrent load. The tests involved concurrent users requesting a term of the Fibonacci sequence (the 10th, 20th, 30th) and the content of a table from the database. The results show that Go achieved the best performance in all benchmark tests. For example, Go reached two times higher throughput than Node.js and five times higher than Apache and Nginx in the 20th Fibonacci term test. In addition, Go had the smallest memory footprint and demonstrated the most efficient resource utilization, in terms of CPU usage. Instead, Node.js had by far the largest memory footprint, consuming up to 90% more memory than Nginx and Apache. Regarding the performance of Apache and Nginx, our findings indicate that Hypertext Preprocessor (PHP) becomes a bottleneck when the servers are requested to respond by performing CPU-intensive tasks under increasing concurrent load.

Keywords: apache, Go, Nginx, node.js, web server benchmarking

Procedia PDF Downloads 97

Authors: Amin Nezarat, Naeime Seifadini

Abstract:

Trip mode selection is a behavioral characteristic of passengers with immense importance for travel demand analysis, transportation planning, and traffic management. Identification of trip mode distribution will allow transportation authorities to adopt appropriate strategies to reduce travel time, traffic and air pollution. The majority of existing trip mode inference models operate based on human selected features and traditional machine learning algorithms. However, human selected features are sensitive to changes in traffic and environmental conditions and susceptible to personal biases, which can make them inefficient. One way to overcome these problems is to use neural networks capable of extracting high-level features from raw input. In this study, the convolutional neural network (CNN) architecture is used to predict the trip mode distribution based on raw GPS trajectory data. The key innovation of this paper is the design of the layout of the input layer of CNN as well as normalization operation, in a way that is not only compatible with the CNN architecture but can also represent the fundamental features of motion including speed, acceleration, jerk, and Bearing rate. The highest prediction accuracy achieved with the proposed configuration for the convolutional neural network with batch normalization is 85.26%.

Keywords: predicting, deep learning, neural network, urban trip

Procedia PDF Downloads 138

Authors: Nor Zaidi Haron, Fauziyah Salehuddin, Norsuhaidah Arshad, Sani Irwan Salim

Abstract:

Resistive random access memories (RRAMs) are one of the main candidates for future computer memories. However, due to their tiny size and immature device technology, the quality of the outgoing RRAM chips is seen as a serious issue. Defective RRAM cells might behave differently than existing semiconductor memories (Dynamic RAM, Static RAM, and Flash), meaning that they are difficult to be detected using existing test schemes. This paper presents an enhanced test scheme, referred to as Programmable Short Write Time (PSWT) that is able to improve the detection of faulty RRAM cells. It is developed by applying multiple weak write operations, each with different time durations. The test circuit embedded in the RRAM chip is made programmable in order to supply different weak write times during testing. The RRAM electrical model is described using Verilog-AMS language and is simulated using HSPICE simulation tools. Simulation results show that the proposed test scheme offers better open-resistive fault detection compared to existing test schemes.

Keywords: memory fault, memory test, design-for-testability, resistive random access memory

Procedia PDF Downloads 387

Authors: Innocent Uzougbo Onwuegbuzie

Abstract:

Wireless Sensor Networks (WSNs) are crucial for numerous applications, yet they face significant challenges due to resource constraints such as limited power and memory. Traditional routing algorithms like Dijkstra, Ad hoc On-Demand Distance Vector (AODV), and Bellman-Ford, while effective in path establishment and discovery, are not optimized for the unique demands of WSNs due to their large memory footprint and power consumption. This paper introduces the Adaptive Energy-Aware Routing (AEAR) model, a solution designed to address these limitations. AEAR integrates reactive route discovery, localized decision-making using geographic information, energy-aware metrics, and dynamic adaptation to provide a robust and efficient routing strategy. We present a detailed comparative analysis using a dataset of 50 sensor nodes, evaluating power consumption, memory footprint, and path cost across AEAR, Dijkstra, AODV, and Bellman-Ford algorithms. Our results demonstrate that AEAR significantly reduces power consumption and memory usage while optimizing path weight. This improvement is achieved through adaptive mechanisms that balance energy efficiency and link quality, ensuring prolonged network lifespan and reliable communication. The AEAR model's superior performance underlines its potential as a viable routing solution for energy-constrained WSN environments, paving the way for more sustainable and resilient sensor network deployments.

Keywords: wireless sensor networks (WSNs), adaptive energy-aware routing (AEAR), routing algorithms, energy, efficiency, network lifespan

Procedia PDF Downloads 35

Authors: A. Alli-Oluwafuyi, A. Amin, S. M. Fii, S. O. Amusa, A. Imam, N. T. Asogwa, W. I. Abdulmajeed, F. Olaseinde, B. V. Owoyele

Abstract:

Chronic stress or prolonged glucocorticoid administration impairs higher cognitive functions in rodents and humans. However, the mechanisms are not fully clear. Insulin and receptors are expressed in the brain and are involved in cognition. Insulin resistance accompanies Alzheimer’s disease and associated cognitive decline. The goal of this study was to evaluate the effects of sub-chronic administration of a glucocorticoid, dexamethasone (DEX) on behavior and biochemical changes in prefrontal cortex (PFC). Male Wistar rats were administered DEX (2, 4 & 8 mg/kg, IP) or saline for seven consecutive days and behavior was assessed in the following paradigms: “Y” maze, elevated plus maze, Morris’ water maze and novel object recognition (NOR) tests. Insulin, lactate dehydrogenase (LDH) and Superoxide Dismutase (SOD) activity were evaluated in homogenates of the prefrontal cortex. DEX-treated rats exhibited impaired prefrontal cortex function manifesting as reduced locomotion, impaired novel object exploration and impaired short- and long-term spatial memory compared to normal controls (p < 0.05). These effects were not consistently dose-dependent. These behavioral alterations were accompanied by a decrease in insulin concentration observed in PFC of 4 mg/kg DEX-treated rats compared to control (10μIU/mg vs. 50μIU/mg; p < 0.05) but not 2mg/kg. Furthermore, we report a modification of brain stress markers LDH and SOD (p > 0.05). These results indicate that prolonged activation of GCs disrupt prefrontal cortex function which may be related to insulin impairment. These effects may not be attributable to a non-specific elevation of oxidative stress in the brain. Future studies would evaluate mechanisms of GR-induced insulin loss.

Keywords: dexamethasone, insulin, memory, prefrontal cortex

Procedia PDF Downloads 284

Authors: Shuen-Tai Wang, Fang-An Kuo, Chau-Yi Chou, Yu-Bin Fang

Abstract:

2016 has become the year of the Artificial Intelligence explosion. AI technologies are getting more and more matured that most world well-known tech giants are making large investment to increase the capabilities in AI. Machine learning is the science of getting computers to act without being explicitly programmed, and deep learning is a subset of machine learning that uses deep neural network to train a machine to learn  features directly from data. Deep learning realizes many machine learning applications which expand the field of AI. At the present time, deep learning frameworks have been widely deployed on servers for deep learning applications in both academia and industry. In training deep neural networks, there are many standard processes or algorithms, but the performance of different frameworks might be different. In this paper we evaluate the running performance of two state-of-the-art distributed deep learning frameworks that are running training calculation in parallel over multi GPU and multi nodes in our cloud environment. We evaluate the training performance of the frameworks with ResNet-50 convolutional neural network, and we analyze what factors that result in the performance among both distributed frameworks as well. Through the experimental analysis, we identify the overheads which could be further optimized. The main contribution is that the evaluation results provide further optimization directions in both performance tuning and algorithmic design.

Keywords: artificial intelligence, machine learning, deep learning, convolutional neural networks

Procedia PDF Downloads 211

Authors: Farzaneh Badinlou, Reza Kormi-Nouri, Monika Knopf

Abstract:

Enacted tasks are typically remembered better than when the same task materials are only verbally encoded, a robust finding referred to as the enactment effect. It has been assumed that enactment effect is independent of object presence but the size of enactment effect can be increased by providing objects at study phase in adults. To clarify the issues in children, free recall and cued recall performance of action phrases with or without using real objects were compared in 410 school-aged children from four age groups (8, 10, 12 and 14 years old). In this study, subjects were instructed to learn a series of action phrases under three encoding conditions, participants listened to verbal action phrases (VTs), performed the phrases (SPTs: subject-performed tasks), and observed the experimenter perform the phrases (EPTs: experimenter-performed tasks). Then, free recall and cued recall memory tests were administrated. The results revealed that the real object compared with imaginary objects improved recall performance in SPTs and EPTs, but more so in VTs. It was also found that the object presence was not necessary for the occurrence of the enactment effect but it was changed the size of enactment effect in all age groups. The size of enactment effect was more pronounced for imaginary objects than the real object in both free recall and cued recall memory tests in children. It was discussed that SPTs and EPTs deferentially facilitate item-specific and relation information processing and providing the objects can moderate the processing underlying the encoding conditions.

Keywords: action memory, enactment effect, item-specific processing, object, relational processing, school-aged children

Procedia PDF Downloads 238

Authors: Teguh Sugiarto, Ahmad Subagyo, Ludiro Madu, Amir Mohammadian Amiri

Abstract:

Recently, the significant fluctuations accosiated with Indonesian economy justifies the need for paying more attention to this issue. In this regard, the main objective of this study is to investigate the relationship between two issues related to the macro Indonesia economy called consumption and GDP during the period of 1967 to 2014. This research method exploits short term and long term relationships using Granger and subsequently, models them by the causality method . However, using analysis of Granger with Johansen shows that there is not only a long term, but also a short-long relationship between GDP and consumption using lags the interval 5.

Keywords: cointegration, Granger causality, GDP, consumption

Procedia PDF Downloads 356

Authors: Yousef Daneshvar Rouyandozagh, Ece. K. Açikgöz

Abstract:

Conservation of historical urban patterns in the traditional neighborhoods is a part of creating integrated urban environments that are socially more sustainable. Urbanization reflects on life conditions and social, physical, economical characteristics of the society. In this regard, historical zones and traditional regions are affected by dramatic interventions on these characteristics. This article focuses on the Uch Dukkan neighborhood located in Ardabil City in Azarbaijani region of Iran, which has been up to such interventions that leaded its transformation from the past to the present. After introducing a brief inventory of the main elements of the historical zone and the neighborhood; this study explores the changes and transformations in different periods; and their impacts on the quality of the environment and its social sustainability. The survey conducted in the neighborhood as part of this research study revealed that the Uch Dukkan neighborhood and the unique architectural heritage that it possesses have become more inactive physically and functionally in a decade. This condition requires an exploration and comparison of the present and the expected transformations of the meaning of social space from the most private unit to the urban scale. From this token, it is argued that an architectural point of view that is based on space order; use and meaning of space as a social and cultural image, should not be ignored. Based on the interplay between social sustainability, collective memory, and the urban environment, study aims to make the invisible portion of ignorance clear, that ends up with a weakness in defining the collective meaning of the neighborhood as a historic urban district. It reveals that the spatial possessions of the neighborhood are valuable not only for their historical and physical characteristics, but also for their social memory that is to be remembered and constructed further.

Keywords: urban integrity, social sustainability, collective memory, social decay

Procedia PDF Downloads 288

Authors: Reggie Chandra

Abstract:

Trafficaccidents and traffic signal optimization are correlated. However, 70-90% of the traffic signals across the USA are not synchronized. The reason behind that is insufficient resources to create and implement timing plans. In this work, we will discuss the use of a breakthrough Artificial Intelligence (AI) technology to optimize traffic flow and collect 24/7/365 accurate traffic data using a vehicle detection system. We will discuss what are recent advances in Artificial Intelligence technology, how does AI work in vehicles, pedestrians, and bike data collection, creating timing plans, and what is the best workflow for that. Apart from that, this paper will showcase how Artificial Intelligence makes signal timing affordable. We will introduce a technology that uses Convolutional Neural Networks (CNN) and deep learning algorithms to detect, collect data, develop timing plans and deploy them in the field. Convolutional Neural Networks are a class of deep learning networks inspired by the biological processes in the visual cortex. A neural net is modeled after the human brain. It consists of millions of densely connected processing nodes. It is a form of machine learning where the neural net learns to recognize vehicles through training - which is called Deep Learning. The well-trained algorithm overcomes most of the issues faced by other detection methods and provides nearly 100% traffic data accuracy. Through this continuous learning-based method, we can constantly update traffic patterns, generate an unlimited number of timing plans and thus improve vehicle flow. Convolutional Neural Networks not only outperform other detection algorithms but also, in cases such as classifying objects into fine-grained categories, outperform humans. Safety is of primary importance to traffic professionals, but they don't have the studies or data to support their decisions. Currently, one-third of transportation agencies do not collect pedestrian and bike data. We will discuss how the use of Artificial Intelligence for data collection can help reduce pedestrian fatalities and enhance the safety of all vulnerable road users. Moreover, it provides traffic engineers with tools that allow them to unleash their potential, instead of dealing with constant complaints, a snapshot of limited handpicked data, dealing with multiple systems requiring additional work for adaptation. The methodologies used and proposed in the research contain a camera model identification method based on deep Convolutional Neural Networks. The proposed application was evaluated on our data sets acquired through a variety of daily real-world road conditions and compared with the performance of the commonly used methods requiring data collection by counting, evaluating, and adapting it, and running it through well-established algorithms, and then deploying it to the field. This work explores themes such as how technologies powered by Artificial Intelligence can benefit your community and how to translate the complex and often overwhelming benefits into a language accessible to elected officials, community leaders, and the public. Exploring such topics empowers citizens with insider knowledge about the potential of better traffic technology to save lives and improve communities. The synergies that Artificial Intelligence brings to traffic signal control and data collection are unsurpassed.

Keywords: artificial intelligence, convolutional neural networks, data collection, signal control, traffic signal

Procedia PDF Downloads 169

Authors: Qian Liu, Steve Furber

Abstract:

To explore how the brain may recognize objects in its general,accurate and energy-efficient manner, this paper proposes the use of a neuromorphic hardware system formed from a Dynamic Video Sensor~(DVS) silicon retina in concert with the SpiNNaker real-time Spiking Neural Network~(SNN) simulator. As a first step in the exploration on this platform a recognition system for dynamic hand postures is developed, enabling the study of the methods used in the visual pathways of the brain. Inspired by the behaviours of the primary visual cortex, Convolutional Neural Networks (CNNs) are modeled using both linear perceptrons and spiking Leaky Integrate-and-Fire (LIF) neurons. In this study's largest configuration using these approaches, a network of 74,210 neurons and 15,216,512 synapses is created and operated in real-time using 290 SpiNNaker processor cores in parallel and with 93.0% accuracy. A smaller network using only 1/10th of the resources is also created, again operating in real-time, and it is able to recognize the postures with an accuracy of around 86.4% -only 6.6% lower than the much larger system. The recognition rate of the smaller network developed on this neuromorphic system is sufficient for a successful hand posture recognition system, and demonstrates a much-improved cost to performance trade-off in its approach.

Keywords: spiking neural network (SNN), convolutional neural network (CNN), posture recognition, neuromorphic system

Procedia PDF Downloads 472

Authors: M. M. Essa, S. Subash, M. Akbar, S. Al-Adawi, A. Al-Asmi, G. J. Guillemein

Abstract:

Transgenic (tg) mice which contain an amyloid precursor protein (APP) gene mutation, develop extracellular amyloid beta (Aβ) deposition in the brain, and severe memory and behavioral deficits with age. These mice serve as an important animal model for testing the efficacy of novel drug candidates for the treatment and management of symptoms of Alzheimer's disease (AD). Several reports have suggested that oxidative stress is the underlying cause of Aβ neurotoxicity in AD. Date palm fruits contain very high levels of antioxidants and several medicinal properties that may be useful for improving the quality of life in AD patients. In this study, we investigated the effect of dietary supplementation of Omani date palm fruits on the memory, anxiety and learning skills along with inflammation in an AD mouse model containing the double Swedish APP mutation (APPsw/Tg2576). The experimental groups of APP-transgenic mice from the age of 4 months were fed custom-mix diets (pellets) containing 2% and 4% Date palm fruits. We assessed spatial memory and learning ability, psychomotor coordination, and anxiety-related behavior in Tg and wild-type mice at the age of 4-5 months and 18-19 months using the Morris water maze test, rota rod test, elevated plus maze test, and open field test. Further, inflammatory parameters also analyzed. APPsw/Tg2576 mice that were fed a standard chow diet without dates showed significant memory deficits, increased anxiety-related behavior, and severe impairment in spatial learning ability, position discrimination learning ability and motor coordination along with increased inflammation compared to the wild type mice on the same diet, at the age of 18-19 months In contrast, PPsw/Tg2576 mice that were fed a diet containing 2% and 4% dates showed a significant improvements in memory, learning, locomotor function, and anxiety with reduced inflammatory markers compared to APPsw/Tg2576 mice fed the standard chow diet. Our results suggest that dietary supplementation with dates may slow the progression of cognitive and behavioral impairments in AD. The exact mechanism is still unclear and further extensive research needed.

Keywords: Alzheimer's disease, date palm fruits, Oman, cognitive decline, memory loss, anxiety, inflammation

Procedia PDF Downloads 423

Authors: sulmaz Mozaffari

Abstract:

Nostalgia is one of the hot-debated issues in critical psychology which has been translated as the yearning or gloom in Persian. It is defined as the regret of the sweet past and the contrast of the present with the past. The feeling of alienation and being remote from the home, remembering death, the regret of childhood and youth, separation of the beloved, remembering the glorious era of history, desire for the ancient times, and the hope for Utopia are considered as its components. Nazik Al-Malaika, a contemporary poet of Arabic literature, has depicted some shapes and dimensions of sympathy, regret and anguish in her poems. Utilizing a nostalgic approach to the past, this paper has reflected upon love, memories of childhood and youth and hope for Utopia "and also aimed at explaining each one's manifestations through a comparative perspective.

Keywords: Nazik al-malaika, poem, nostalgia, personal memory, collective memory

Procedia PDF Downloads 410

Authors: Donatello Cardone

Abstract:

Shape memory alloys (SMAs) have great potential for use in the field of civil engineering. The author of this manuscript has been involved, since 1996, in several experimental and theoretical studies on the application of SMAs in structural engineering, within national and international research projects. This paper provides an overview of the main results achieved, including the conceptual design, implementation, and testing of different SMA-based devices, namely: (i) energy-dissipating braces for RC buildings, (ii) seismic isolation devices for buildings and bridges, (iii) smart tie-rods for arches and vaults and (iv) seismic restrainers for bridges. The main advantages of using SMA-based devices in the seismic protection of structures derive from the double-flag shape of their hysteresis loops, which implies three favourable features, i.e., self-centering capability, good energy dissipation capability, and high stiffness for small displacements. The main advantages of SMA-based units for steel tie-rods are associated with the thermal behaviour of superelastic SMAs, which is antagonistic compared to that of steel. This implies a strong reduction of force changes due to air temperature variations. Finally, SMA-based seismic restrainers proved to be effective in preventing bridge deck unseating and pounding.

Keywords: seismic protection of structures, shape memory alloys, structural engineering, steel tie-rods, seismic restrainers for bridges

Procedia PDF Downloads 97

Authors: Bipasha Sen, Aditya Agarwal

Abstract:

Multilingual automatic speech recognition (ASR) system is a single entity capable of transcribing multiple languages sharing a common phone space. Performance of such a system is highly dependent on the compatibility of the languages. State of the art speech recognition systems are built using sequential architectures based on recurrent neural networks (RNN) limiting the computational parallelization in training. This poses a significant challenge in terms of time taken to bootstrap and validate the compatibility of multiple languages for building a robust multilingual system. Complex architectural choices based on self-attention networks are made to improve the parallelization thereby reducing the training time. In this work, we propose Reed, a simple system based on 1D convolutions which uses very short context to improve the training time. To improve the performance of our system, we use raw time-domain speech signals directly as input. This enables the convolutional layers to learn feature representations rather than relying on handcrafted features such as MFCC. We report improvement on training and inference times by atleast a factor of 4x and 7.4x respectively with comparable WERs against standard RNN based baseline systems on SpeechOcean's multilingual low resource dataset.

Keywords: convolutional neural networks, language compatibility, low resource languages, multilingual automatic speech recognition

Procedia PDF Downloads 123

Authors: Aswathi Thrivikraman, S. Advaith

Abstract:

The study is aimed at designing encoders for deciphering intricacies in time series data by redescribing the dynamics operating on a lower-dimensional manifold. A semi-supervised LSTM autoencoder is devised and investigated to see if the latent representation of the time series data can better forecast the data. End-to-end training of the LSTM autoencoder, together with another LSTM network that is connected to the latent space, forces the hidden states of the encoder to represent the most meaningful latent variables relevant for forecasting. Furthermore, the study compares the predictions with those of a traditional seq2seq model.

Keywords: LSTM, autoencoder, forecasting, seq2seq model

Procedia PDF Downloads 155

Authors: Arezou Hajhashemi, Karim Asgari, Masoud Etemadifar, Maryam Keyvani, Ali Hekmatnia

Abstract:

Multiple Sclerosis (MS) is one of the most prevalent demyelinating diseases in CNS. As in other chronic cerebral diseases, impairment in cognitive functioning and in memory is popular. Because of the inflammatory and demyelinating nature of the disease, the localization of plaques in different parts of the Prefrontal and Limbic System, may lead to memorial symptoms. This investigation was intended to study relationship between frequency of plaques and memorial symptoms arising from dysfunction limbic system and prefrontal of patients with MS. The sample was selected randomly from patients with MS with memory problem, who have been referred to Isfahan Multiple Sclerosis Society. Brain System Test and Memory Test was administered to the sample, and their MRI's were analyzed by specialist in order to indentify two different parts of plaques. The data was analyzed by SPSS. The results showed that there were significant relationship between MS plaques and prefrontal's dysfunction and memorial symptom related to prefrontal area; however, there were no significant relationship between MS plaques and limbic system's dysfunction and memorial symptoms related to limbic system area. The results of this study suggest that memorial symptoms due to injury regions of the brain have the most significant relationship to prefrontal. Better judgment about these results needs more studies in future.

Keywords: multiple sclerosis, magnetic image, brain injury, behavior disorder

Procedia PDF Downloads 514

Authors: Evan Lowhorn, Rocio Alba-Flores

Abstract:

The purpose of this work is to develop a method for classifying hand signals and using the output in a drone control algorithm. To achieve this, methods based on Convolutional Neural Networks (CNN) were applied. CNN's are a subset of deep learning, which allows grid-like inputs to be processed and passed through a neural network to be trained for classification. This type of neural network allows for classification via imaging, which is less intrusive than previous methods using biosensors, such as EMG sensors. Classification CNN's operate purely from the pixel values in an image; therefore they can be used without additional exteroceptive sensors. A development bench was constructed using a desktop computer connected to a high-definition webcam mounted on a scissor arm. This allowed the camera to be pointed downwards at the desk to provide a constant solid background for the dataset and a clear detection area for the user. A MATLAB script was created to automate dataset image capture at the development bench and save the images to the desktop. This allowed the user to create their own dataset of 12,000 images within three hours. These images were evenly distributed among seven classes. The defined classes include forward, backward, left, right, idle, and land. The drone has a popular flip function which was also included as an additional class. To simplify control, the corresponding hand signals chosen were the numerical hand signs for one through five for movements, a fist for land, and the universal “ok” sign for the flip command. Transfer learning with PyTorch (Python) was performed using a pre-trained 18-layer residual learning network (ResNet-18) to retrain the network for custom classification. An algorithm was created to interpret the classification and send encoded messages to a Ryze Tello drone over its 2.4 GHz Wi-Fi connection. The drone’s movements were performed in half-meter distance increments at a constant speed. When combined with the drone control algorithm, the classification performed as desired with negligible latency when compared to the delay in the drone’s movement commands.

Keywords: classification, computer vision, convolutional neural networks, drone control

Procedia PDF Downloads 210

Authors: Thaveesha Dheerasekera, Dileeka Sandamali Alwis

Abstract:

The escalating prevalence of mental health issues, such as depression and suicidal ideation, is a matter of significant global concern. It is plausible that a variety of factors, such as life events, social isolation, and preexisting physiological or psychological health conditions, could instigate or exacerbate these conditions. Traditional approaches to diagnosing depression entail a considerable amount of time and necessitate the involvement of adept practitioners. This underscores the necessity for automated systems capable of promptly detecting and diagnosing symptoms of depression. The PsyVBot system employs sophisticated natural language processing and machine learning methodologies, including the use of the NLTK toolkit for dataset preprocessing and the utilization of a Long Short-Term Memory (LSTM) model. The PsyVBot exhibits a remarkable ability to diagnose depression with a 94% accuracy rate through the analysis of user input. Consequently, this resource proves to be efficacious for individuals, particularly those enrolled in academic institutions, who may encounter challenges pertaining to their psychological well-being. The PsyVBot employs a Long Short-Term Memory (LSTM) model that comprises a total of three layers, namely an embedding layer, an LSTM layer, and a dense layer. The stratification of these layers facilitates a precise examination of linguistic patterns that are associated with the condition of depression. The PsyVBot has the capability to accurately assess an individual's level of depression through the identification of linguistic and contextual cues. The task is achieved via a rigorous training regimen, which is executed by utilizing a dataset comprising information sourced from the subreddit r/SuicideWatch. The diverse data present in the dataset ensures precise and delicate identification of symptoms linked with depression, thereby guaranteeing accuracy. PsyVBot not only possesses diagnostic capabilities but also enhances the user experience through the utilization of audio outputs. This feature enables users to engage in more captivating and interactive interactions. The PsyVBot platform offers individuals the opportunity to conveniently diagnose mental health challenges through a confidential and user-friendly interface. Regarding the advancement of PsyVBot, maintaining user confidentiality and upholding ethical principles are of paramount significance. It is imperative to note that diligent efforts are undertaken to adhere to ethical standards, thereby safeguarding the confidentiality of user information and ensuring its security. Moreover, the chatbot fosters a conducive atmosphere that is supportive and compassionate, thereby promoting psychological welfare. In brief, PsyVBot is an automated conversational agent that utilizes an LSTM model to assess the level of depression in accordance with the input provided by the user. The demonstrated accuracy rate of 94% serves as a promising indication of the potential efficacy of employing natural language processing and machine learning techniques in tackling challenges associated with mental health. The reliability of PsyVBot is further improved by the fact that it makes use of the Reddit dataset and incorporates Natural Language Toolkit (NLTK) for preprocessing. PsyVBot represents a pioneering and user-centric solution that furnishes an easily accessible and confidential medium for seeking assistance. The present platform is offered as a modality to tackle the pervasive issue of depression and the contemplation of suicide.

Keywords: chatbot, depression diagnosis, LSTM model, natural language process

Procedia PDF Downloads 68

Authors: Gonzalo Reyes, Magdalena Walczak, Esteban Ramos-Moore, Jorge Ramos-Grez

Abstract:

Metal additive manufacture technologies have gained strong support and acceptance as a promising and alternative method to manufacture high performance complex geometry products. The main purpose of the present work is to study the microstructure and phase transformation temperatures of Cu-Al-Ni shape memory alloys fabricated from a direct laser additive process using metallic powders as precursors. The potential application is to manufacture self-centering seismic dampers for earthquake protection of buildings out of a copper based alloy by an additive process. In this process, the Cu-Al-Ni alloy is melted, inside of a high temperature and vacuum chamber with the aid of a high power fiber laser under inert atmosphere. The laser provides the energy to melt the alloy powder layer. The process allows fabricating fully dense, oxygen-free Cu-Al-Ni specimens using different laser power levels, laser powder interaction times, furnace ambient temperatures, and cooling rates as well as modifying concentration of the alloying elements. Two sets of specimens were fabricated with a nominal composition of Cu-13Al-3Ni and Cu-13Al-4Ni in wt.%, however, semi-quantitative chemical analysis using EDX examination showed that the specimens’ resulting composition was closer to Cu-12Al-5Ni and Cu-11Al-8Ni, respectively. In spite of that fact, it is expected that the specimens should still possess shape memory behavior. To confirm this hypothesis, phase transformation temperatures will be measured using DSC technique, to look for martensitic and austenitic phase transformations at 150°C. So far, metallographic analysis of the specimens showed defined martensitic microstructures. Moreover, XRD technique revealed diffraction peaks corresponding to (0 0 18) and (1 2 8) planes, which are too associated with the presence of martensitic phase. We conclude that it would be possible to obtain fully dense Cu-Al-Ni alloys having shape memory effect behavior by direct laser fabrication process, and to advance into fabrication of self centering seismic dampers by a controllable metal additive manufacturing process.

Keywords: Cu-Al-Ni alloys, direct laser fabrication, shape memory alloy, self-centering seismic dampers

Procedia PDF Downloads 516

Authors: Soumyajit Koley, Kuladeep Rajamudili, Supriyo Ganguly

Abstract:

In recent years, iron-based shape-memory alloys have been emerging as an inexpensive alternative to costly Ni-Ti alloy and thus considered suitable for many different applications in civil structures. Fe-17Mn-10Cr-5Si-4Ni-0.5V-0.5C alloy contains 37 wt.% of total solute elements. Such complex multi-component metallurgical system often leads to severe solute segregation and solidification cracking. Wire-arc additive manufacturing (WAAM) of Fe-17Mn-10Cr-5Si-4Ni-0.5V-0.5C alloy was attempted using a cold-wire fed plasma arc torch attached to a 6-axis robot. Self-standing walls were manufactured. However, multiple vertical cracks were observed after deposition of around 15 layers. Microstructural characterization revealed open surfaces of dendrites inside the crack, confirming these cracks as solidification cracks. Machine hammer peening (MHP) process was adopted on each layer to cold work the newly deposited alloy. Effect of MHP traverse speed were varied systematically to attain a window of operation where cracking was completely stopped. Microstructural and textural analysis were carried out further to correlate the peening process to microstructure.MHP helped in many ways. Firstly, a compressive residual stress was induced on each layer which countered the tensile residual stress evolved from solidification process; thus, reducing net tensile stress on the wall along its length. Secondly, significant local plastic deformation from MHP followed by the thermal cycle induced by deposition of next layer resulted into a recovered and recrystallized equiaxed microstructure instead of long columnar grains along the vertical direction. This microstructural change increased the total crack propagation length and thus, the overall toughness. Thirdly, the inter-layer peening significantly reduced the strong cubic {001} crystallographic texture formed along the build direction. Cubic {001} texture promotes easy separation of planes and easy crack propagation. Thus reduction of cubic texture alleviates the chance of cracking.

Keywords: Iron-based shape-memory alloy, wire-arc additive manufacturing, solidification cracking, inter-layer cold working, machine hammer peening

Procedia PDF Downloads 72

Authors: Lai Gwen Chan, Yining Ong, Audrey Yoke Poh Wong

Abstract:

Background. The efficacy of electroconvulsive therapy (ECT) as a form of treatment to a range of mental disorders is well-established. However, ECT is often associated with either temporary or persistent cognitive side-effects, resulting in the failure of wider prescription. Of which, retrograde amnesia is the most commonly reported cognitive side-effect. Most studies found a recalling deficit in autobiographical memories to be short-term, although a few have reported more persistent amnesic effects. Little is known about ECT-related amnesic effects in Asian population. Hence, this study aims to resolve conflicting findings, as well as to better elucidate the effects of ECT on cognitive functioning in a local sample. Method: 12 patients underwent bilateral ECT under the care of Psychological Medicine Department, Tan Tock Seng Hospital, Singapore. Participants’ cognition and level of functioning were assessed at four time-points: before ECT, between the third and fourth induced seizure, at the end of the whole course of ECT, and two months after the index course of ECT. Results: It was found that Global Assessment of Functioning scores increased significantly at the completion of ECT. Case-by-case analyses also revealed an overall improvement in Personal Semantic and Autobiographical memory two months after the index course of ECT. A transient dip in both personal semantic and autobiographical memory scores was observed in one participant between the third and fourth induced seizure, but subsequently resolved and showed better performance than at baseline. Conclusions: The findings of this study suggest that ECT is an effective form of treatment to alleviate the severity of symptoms of the diagnosis. ECT does not affect attention, language, executive functioning, personal semantic and autobiographical memory adversely. The findings suggest that Asian patients may respond to bilateral ECT differently from Western samples.

Keywords: electroconvulsive therapy (ECT), autobiographical memory, cognitive impairment, psychiatric disorder

Procedia PDF Downloads 193

Authors: Farzad Hosseini Hossein Abadi, Cristina Prieto Sierra, Cesar Álvarez Díaz

Abstract:

Predicting streamflow accurately in flashy catchments prone to floods is a major research and operational challenge in hydrological modeling. Recent advancements in deep learning, particularly Long Short-Term Memory (LSTM) networks, have shown to be promising in achieving accurate hydrological predictions at daily and hourly time scales. In this work, a multi-timescale LSTM (MTS-LSTM) network was applied to the context of regional hydrological predictions at an hourly time scale in flashy catchments. The case study includes 40 catchments allocated in the Basque Country, north of Spain. We explore the impact of hyperparameters on the performance of streamflow predictions given by regional deep learning models through systematic hyperparameter tuning - where optimal regional values for different catchments are identified. The results show that predictions are highly accurate, with Nash-Sutcliffe (NSE) and Kling-Gupta (KGE) metrics values as high as 0.98 and 0.97, respectively. A principal component analysis reveals that a hyperparameter related to the length of the input sequence contributes most significantly to the prediction performance. The findings suggest that input sequence lengths have a crucial impact on the model prediction performance. Moreover, employing catchment-scale analysis reveals distinct sequence lengths for individual basins, highlighting the necessity of customizing this hyperparameter based on each catchment’s characteristics. This aligns with well known “uniqueness of the place” paradigm. In prior research, tuning the length of the input sequence of LSTMs has received limited focus in the field of streamflow prediction. Initially it was set to 365 days to capture a full annual water cycle. Later, performing limited systematic hyper-tuning using grid search, revealed a modification to 270 days. However, despite the significance of this hyperparameter in hydrological predictions, usually studies have overlooked its tuning and fixed it to 365 days. This study, employing a simultaneous systematic hyperparameter tuning approach, emphasizes the critical role of input sequence length as an influential hyperparameter in configuring LSTMs for regional streamflow prediction. Proper tuning of this hyperparameter is essential for achieving accurate hourly predictions using deep learning models.

Keywords: LSTMs, streamflow, hyperparameters, hydrology

Procedia PDF Downloads 69