Search results for: convolutional long short-term memory

Authors: Kang Huang, Wanting Zhou, Shiwei Yuan, Lei Li

Abstract:

Since information technology develops rapidly, the security issue has become an increasingly critical for computer system. In particular, as cloud computing and the Internet of Things (IoT) continue to gain widespread adoption, computer systems need to new security threats and attacks. The Root of Trust (RoT) is the foundation for providing basic trusted computing, which is used to verify the security and trustworthiness of other components. Design a reliable Root of Trust and guarantee its own security are essential for improving the overall security and credibility of computer systems. In this paper, we discuss the implementation of self-security technology based on the RISC-V Root of Trust at the hardware level. To effectively safeguard the security of the Root of Trust, researches on security safeguard technology on the Root of Trust have been studied. At first, a lightweight and secure boot framework is proposed as a secure mechanism. Secondly, two kinds of memory protection mechanism are built to against memory attacks. Moreover, hardware implementation of proposed method has been also investigated. A series of experiments and tests have been carried on to verify to effectiveness of the proposed method. The experimental results demonstrated that the proposed approach is effective in verifying the integrity of the Root of Trust’s own boot rom, user instructions, and data, ensuring authenticity and enabling the secure boot of the Root of Trust’s own system. Additionally, our approach provides memory protection against certain types of memory attacks, such as cache leaks and tampering, and ensures the security of root-of-trust sensitive information, including keys.

Keywords: root of trust, secure boot, memory protection, hardware security

Procedia PDF Downloads 216

Authors: Iliyasu Musa Omoyine, Musa Sunday Abraham, Oladele Sunday Blessing, Iliya Ibrahim Abdullahi, Ibegbu Augustine Oseloka, Nuhu Nana-Hawau, Animoku Abdulrazaq Amoto, Yusuf Abdullateef Onoruoiza, Sambo Sohnap James, Akpulu Steven Peter, Ajayi Abayomi

Abstract:

The neuroprotective role of inflammation from detrimental intrinsic and extrinsic factors has been reported. However, the overactivation of astrocytes and microglia due to lead toxicity produce excessive pro-inflammatory cytokines, mediating neurodegenerative diseases. The present study investigated the mitigatory effects of quercetin on neuroinflammation, correlating with memory function in lead-exposed rats. In this study, Wistar rats were administered orally with Quercetin (Q: 60 mg/kg) and Succimer as a standard drug (S: 10 mg/kg) for 21 days after lead exposure (Pb: 125 mg/kg) of 21 days or in combination with Pb, once daily for 42 days. Working and reference memory was assessed using an Eight-arm radial water maze (8-ARWM). The changes in brain lead level, the neuronal nitric oxide synthase (nNOS) activity, and the level of neuroinflammatory markers such as tumour necrosis factor-alpha (TNF-α) and Interleukin 1 Beta (IL-1β) were determined. Immunohistochemically, astrocyte expression was evaluated. The results showed that the brain level of lead was increased significantly in lead-exposed rats. The expression of astrocytes increased in the CA3 and CA1 regions of the hippocampus, and the levels of brain TNF-α and IL-1β increased in lead-exposed rats. Lead impaired reference and working memory by increasing reference memory errors and working memory incorrect errors in lead-exposed rats. However, quercetin treatment effectively improved memory and inhibited neuroinflammation by reducing astrocytes’ expression and the levels of TNF-α and IL-1β. The expression of astrocytes and the levels of TNF-α and IL-1β correlated with memory function. The possible explanation for quercetin’s anti-neuroinflammatory effect is that it modulates the activity of cellular proteins involved in the inflammatory response; inhibits the transcription factor of nuclear factor-kappa B (NF-κB), which regulates the expression of proinflammatory molecules; inhibits kinases required for the synthesis of Glial fibrillary acidic protein (GFAP) and modifies the phosphorylation of some proteins, which affect the structure and function of intermediate filament proteins; and, lastly, induces Cyclic-AMP Response Element Binding (CREB) activation and neurogenesis as a compensatory mechanism for memory deficits and neuronal cell death. In conclusion, the levels of neuroinflammatory markers negatively correlated with memory function. Thus, quercetin may be a promising therapy in neuroinflammation and memory dysfunction in populations prone to lead exposure.

Keywords: lead, quercetin, neuroinflammation, memory

Procedia PDF Downloads 54

Authors: Hela Krir, Abdelhak Ayadi, Chedly Bradaii

Abstract:

The influence of both intrinsic factors, elastic energy and memory effect, and radial flow on the appearance and the evolution of the extrudate swelling are investigated in the present work. The experiments have been performed with linear polydimethylsiloxane (PDMS) via a capillary rheometer in which a convergent radial flow was created upstream the contraction. The correspondence between the effects of radial flow, entry elastic stored energy and memory effect is discussed. In particular, as the influence of the considered radial flow, extrudate photographs showed that when the gap ratio is reduced, the extrudate swell is lessened than what it is when radial flow geometry is not installed. Moreover, with a narrower gap, the polymer stores less energy during its passage through the die which implies a lower extrudate swelling at the outlet of the die. Results previously mentioned may be related both to shear and elongational components of radial flow.

Keywords: elastic energy, extrudate swell, memory effect, radial flow

Procedia PDF Downloads 172

Authors: Kang Zijian, Zhang Tingyu, Burra Venkata Durga Kumar

Abstract:

This article investigates the challenges in memory migration during the live migration of virtual machines. We found three challenges probably existing in pre-copy technology. One of the main challenges is the challenge of downtime migration. Decrease the downtime could promise the normal work for a virtual machine. Although pre-copy technology is greatly decreasing the downtime, we still need to shut down the machine in order to finish the last round of data transfer. This paper provides an optimization scheme for the problems existing in pro-copy technology, mainly the optimization of the dirty page migration mechanism. The typical pre-copy technology copy n-1th’s dirty pages in nth turn. However, our idea is to create a double iteration method to solve this problem.

Keywords: virtual machine, pre-copy technology, memory migration process, downtime, dirty pages migration method

Procedia PDF Downloads 151

Authors: Maria Korman, Carmit Gal, Ella Gabitov, Avi Karni

Abstract:

The main question addressed in this study was whether the time-of-day wherein training is afforded is a significant factor for motor skill ('how-to', procedural knowledge) acquisition and consolidation into long term memory in the healthy elderly population. Twenty-nine older adults (60-75 years) practiced an explicitly instructed 5-element key-press sequence by repeatedly generating the sequence ‘as fast and accurately as possible’. Contribution of three parameters to acquisition, 24h post-training consolidation, and 1-week retention gains in motor sequence speed was assessed: (a) time of training (morning vs. evening group) (b) sleep quality (actigraphy) and (c) chronotype. All study participants were moderately morning type, according to the Morningness-Eveningness Questionnaire score. All participants had sleep patterns typical of age, with average sleep efficiency of ~ 82%, and approximately 6 hours of sleep. Speed of motor sequence performance in both groups improved to a similar extent during training session. Nevertheless, evening group expressed small but significant overnight consolidation phase gains, while morning group showed only maintenance of performance level attained at the end of training. By 1-week retention test, both groups showed similar performance levels with no significant gains or losses with respect to 24h test. Changes in the tapping patterns at 24h and 1-week post-training were assessed based on normalized Pearson correlation coefficients using the Fisher’s z-transformation in reference to the tapping pattern attained at the end of the training. Significant differences between the groups were found: the evening group showed larger changes in tapping patterns across the consolidation and retention windows. Our results show that morning-oriented older adults effectively acquired, consolidated, and maintained a new sequence of finger movements, following both morning and evening practice sessions. However, time-of-training affected the time-course of skill evolution in terms of performance speed, as well as the re-organization of tapping patterns during the consolidation period. These results are in line with the notion that motor training preceding a sleep interval may be beneficial for the long-term memory in the elderly. Evening training should be considered an appropriate time window for motor skill learning in older adults, even in individuals with morning chronotype.

Keywords: time-of-day, elderly, motor learning, memory consolidation, chronotype

Procedia PDF Downloads 134

Authors: Lei Zhu, Nan Li

Abstract:

Cold stamping has been widely applied in the automotive industry for the mass production of a great range of automotive panels. Predicting the springback to ensure the dimensional accuracy of the cold-stamped components is a critical step. The main approaches for the prediction and compensation of springback in cold stamping include running Finite Element (FE) simulations and conducting experiments, which require forming process expertise and can be time-consuming and expensive for the design of cold stamping tools. Machine learning technologies have been proven and successfully applied in learning complex system behaviours using presentative samples. These technologies exhibit the promising potential to be used as supporting design tools for metal forming technologies. This study, for the first time, presents a novel application of a Convolutional Neural Network (CNN) based surrogate model to predict the springback fields for variable U-shape cold bending geometries. A dataset is created based on the U-shape cold bending geometries and the corresponding FE simulations results. The dataset is then applied to train the CNN surrogate model. The result shows that the surrogate model can achieve near indistinguishable full-field predictions in real-time when compared with the FE simulation results. The application of CNN in efficient springback prediction can be adopted in industrial settings to aid both conceptual and final component designs for designers without having manufacturing knowledge.

Keywords: springback, cold stamping, convolutional neural networks, machine learning

Procedia PDF Downloads 149

Authors: Shaik Jilani Basha, Inavolu Avinash, Alla Venu Sai Reddy, Bitragunta Taraka Ramu

Abstract:

We can use the neural style transfer technique to build a picture with the same "content" as the beginning image but the "style" of the picture we've chosen. Neural style transfer is a technique for merging the style of one image into another while retaining its original information. The only change is how the image is formatted to give it an additional artistic sense. The content image depicts the plan or drawing, as well as the colors of the drawing or paintings used to portray the style. It is a computer vision programme that learns and processes images through deep convolutional neural networks. To implement software, we used to train deep learning models with the train data, and whenever a user takes an image and a styled image, the output will be as the style gets transferred to the original image, and it will be shown as the output.

Keywords: neural networks, computer vision, deep learning, convolutional neural networks

Procedia PDF Downloads 95

Authors: Bliss Singhal

Abstract:

Epilepsy is a prevalent neurological disorder affecting approximately 50 million individuals worldwide and 1.2 million Americans. There exist millions of pediatric patients with intractable epilepsy, a condition in which seizures fail to come under control. The occurrence of seizures can result in physical injury, disorientation, unconsciousness, and additional symptoms that could impede children's ability to participate in everyday tasks. Predicting seizures can help parents and healthcare providers take precautions, prevent risky situations, and mentally prepare children to minimize anxiety and nervousness associated with the uncertainty of a seizure. This research proposes a comprehensive framework to predict seizures in pediatric patients by evaluating machine learning algorithms on unimodal neuroimaging data consisting of electroencephalogram signals. The bandpass filtering and independent component analysis proved to be effective in reducing the noise and artifacts from the dataset. Various machine learning algorithms’ performance is evaluated on important metrics such as accuracy, precision, specificity, sensitivity, F1 score and MCC. The results show that the deep learning algorithms are more successful in predicting seizures than logistic Regression, and k nearest neighbors. The recurrent neural network (RNN) gave the highest precision and F1 Score, long short-term memory (LSTM) outperformed RNN in accuracy and convolutional neural network (CNN) resulted in the highest Specificity. This research has significant implications for healthcare providers in proactively managing seizure occurrence in pediatric patients, potentially transforming clinical practices, and improving pediatric care.

Keywords: intractable epilepsy, seizure, deep learning, prediction, electroencephalogram channels

Procedia PDF Downloads 84

Authors: Abdul Wahid

Abstract:

Inner speech, also known as verbal thinking, self-talk or private speech, is characterized by the subjective language experience in the absence of overt or audible speech. It is a psychological form of verbal activity which is being rehearsed without the articulation of any sound wave. In Psychology, self-limiting speech means the type of speech which contains information that inhibits the development of the self. People, in most cases, experience inner speech in their first language. It is very frequent in Bangladesh where the Bangla (L1) speaking students lose track of speech during their presentations in English (L2). This paper investigates into the long pauses (more than 0.4 seconds long) in English (L2) presentations by Bangla speaking students (18-21 year old) and finds the intervention of Bangla (L1) inner speech as one of its causes. The overt speeches of the presenters are placed on Audacity Audio Editing software where the length of pauses are measured in milliseconds. Varieties of inner speech questionnaire (VISQ) have been conducted randomly amongst the participants out of whom 20 were selected who have similar phenomenology of inner speech. They have been interviewed to describe the type and content of the voices that went on in their head during the long pauses. The qualitative interview data are then codified and converted into quantitative data. It was observed that in more than 80% cases students experience self-limiting inner speech/self-talk during their unwanted pauses in L2 presentations.

Keywords: Bangla-English Bilinguals, inner speech, L1 intervention in bilingualism, motor schema, pauses, phonological loop, phonological store, working memory

Procedia PDF Downloads 152

Authors: Ingo Roden, Stefana Lupu, Mara Krone, Jasmin Chantah, Gunter Kreutz, Stephan Bongard, Dietmar Grube

Abstract:

The present experimental study examined the effects of music and math training on mathematical skills and visuospatial working memory capacity in kindergarten children. For this purpose, N = 54 children (mean age: 5.46 years; SD = .29) were randomly assigned to three groups. Children in the music group (n = 18) received weekly sessions of 60 min music training over a period of eight weeks, whereas children in the math group (n = 18) received the same amount of training focusing on mathematical basic skills, such as numeracy skills, quantity comparison, and counting objectives. The third group of children (n = 18) served as waiting controls. The groups were matched for sex, age, IQ and previous music experiences at baseline. Pre-Post intervention measurements revealed a significant interaction effect of group x time, showing that children in both music and math groups significantly improved their early numeracy skills, whereas children in the control group did not. No significant differences between groups were observed for the visuospatial working memory performances. These results confirm and extend previous findings on transfer effects of music training on mathematical abilities and visuospatial working memory capacity. They show that music and math interventions are similarly effective to enhance children’s mathematical skills. More research is necessary to establish, whether cognitive transfer effects arising from music interventions might facilitate children’s transition from kindergarten to first-grade.

Keywords: music training, mathematical skills, working memory, transfer

Procedia PDF Downloads 272

Authors: Jérémie Ochin

Abstract:

Soccer analytics are built on two data sources: the frame-by-frame position of each player on the terrain and the sequences of events, such as ball drive, pass, cross, shot, throw-in... With more than 2000 ball-events per soccer game, their precise and exhaustive annotation, based on a monocular video stream such as a TV broadcast, remains a tedious and costly manual task. State-of-the-art methods for spatio-temporal action detection from a monocular video stream, often based on 3D convolutional neural networks, are close to reach levels of performances in mean Average Precision (mAP) compatibles with the automation of such task. Nevertheless, to meet their expectation of exhaustiveness in the context of data analytics, such methods must be applied in a regime of high recall – low precision, using low confidence score thresholds. This setting unavoidably leads to the detection of false positives that are the product of the well documented overconfidence behaviour of neural networks and, in this case, their limited access to contextual information and understanding of the game: their predictions are highly unstructured. Based on the assumption that professional soccer players’ behaviour, pose, positions and velocity are highly interrelated and locally driven by the player performing a ball-action, it is hypothesized that the addition of information regarding surrounding player’s appearance, positions and velocity in the prediction methods can improve their metrics. Several methods are compared to build a proper representation of the game surrounding a player, from handcrafted features of the local graph, based on domain knowledge, to the use of Graph Neural Networks trained in an end-to-end fashion with existing state-of-the-art 3D convolutional neural networks. It is shown that the inclusion of information regarding surrounding players helps reaching higher metrics.

Keywords: fine-grained action recognition, human action recognition, convolutional neural networks, graph neural networks, spatio-temporal action recognition

Procedia PDF Downloads 23

Authors: Jian Li

Abstract:

The uncertainty quantification and inversion problems of subsurface oil-water phase flow usually require extensive repeated forward calculations for new runs with changed conditions. To reduce the computational time, various forms of surrogate models have been built. Related research shows that deep learning has emerged as an effective surrogate model, while most surrogate models with deep learning are purely data-driven, which always leads to poor robustness and abnormal results. To guarantee the model more consistent with the physical laws, a coupled theory-guided convolutional neural network (TgCNN) based surrogate model is built to facilitate computation efficiency under the premise of satisfactory accuracy. The model is a convolutional neural network based on multi-well reservoir simulation. The core notion of this proposed method is to bridge two separate blocks on top of an overall network. They underlie the TgCNN model in a coupled form, which reflects the coupling nature of pressure and water saturation in the two-phase flow equation. The model is driven by not only labeled data but also scientific theories, including governing equations, stochastic parameterization, boundary, and initial conditions, well conditions, and expert knowledge. The results show that the TgCNN-based surrogate model exhibits satisfactory accuracy and efficiency in subsurface oil-water phase flow under multi-well conditions.

Keywords: coupled theory-guided convolutional neural network, multi-well conditions, surrogate model, subsurface oil-water phase

Procedia PDF Downloads 86

Authors: Łukasz Bis

Abstract:

The author is going to discuss virtual reality and its importance for new didactic methods. It has been known for years that experience-based education gives much better results in terms of long-term memory than theoretical study. However, practice is expensive - virtual reality allows the use of an empirical approach to learning, with minimized production costs. The author defines what makes a given VR experience appropriate (adequate) for the didactic and cognitive process. The article is a kind of a list of guidelines and their importance for the VR experience under development.

Keywords: virtual reality, education, universal design, guideline

Procedia PDF Downloads 106

Authors: Katielly Vianna Polkowski, Rodrigo Denizarte de Oliveira Polkowski, Cristiano Grings Herbert

Abstract:

The use of nanomaterials in the formulation of polymeric materials modifies their molecular structure, offering an infinite range of possibilities for the development of smart products, being of great importance for science and contemporary industry. Shape memory polymers are generally lightweight, have high shape recovery capabilities, they are easy to process and have properties that can be adapted for a variety of applications. Shape memory materials are active materials that have attracted attention due to their superior damping properties when compared to conventional structural materials. The development of methodologies capable of preparing new materials, which use graphene in their structure, represents technological innovation that transforms low-cost products into advanced materials with high added value. To obtain an improvement in the shape memory effect (SME) of polymeric materials, it is possible to use graphene in its composition containing low concentration by mass of graphene nanoplatelets (GNP), graphene oxide (GO) or other functionalized graphene, via different mixture process. As a result, there was an improvement in the SME, regarding the increase in the values of maximum strain. In addition, the use of graphene contributes to obtaining nanocomposites with superior electrical properties, greater crystallinity, as well as resistance to material degradation. The methodology used in the research is Systematic Review, scientific investigation, gathering relevant studies on influence of nanomaterials on the properties of shape memory polymeric, using the literature database as a source and study methods. In the present study, a systematic reviewwas performed of all papers published from 2014 to 2022 regarding graphene and shape memory polymeric througha search of three databases. This study allows for easy identification of themost relevant fields of study with respect to graphene and shape memory polymeric, as well as the main gaps to beexplored in the literature. The addition of graphene showed improvements in obtaining higher values of maximum deformation of the material, attributed to a possible slip between stacked or agglomerated nanostructures, as well as an increase in stiffness due to the increase in the degree of phase separation that results in a greater amount physical cross-links, referring to the formation of shortrange rigid domains.

Keywords: graphene, shape memory, smart materials, polymers, nanomaterials

Procedia PDF Downloads 84

Authors: Eiman Kattan, Hong Wei

Abstract:

In using Convolutional Neural Network (CNN) for classification, there is a set of hyperparameters available for the configuration purpose. This study aims to evaluate the impact of a range of parameters in CNN architecture i.e. AlexNet on land cover classification based on four remotely sensed datasets. The evaluation tests the influence of a set of hyperparameters on the classification performance. The parameters concerned are epoch values, batch size, and convolutional filter size against input image size. Thus, a set of experiments were conducted to specify the effectiveness of the selected parameters using two implementing approaches, named pertained and fine-tuned. We first explore the number of epochs under several selected batch size values (32, 64, 128 and 200). The impact of kernel size of convolutional filters (1, 3, 5, 7, 10, 15, 20, 25 and 30) was evaluated against the image size under testing (64, 96, 128, 180 and 224), which gave us insight of the relationship between the size of convolutional filters and image size. To generalise the validation, four remote sensing datasets, AID, RSD, UCMerced and RSCCN, which have different land covers and are publicly available, were used in the experiments. These datasets have a wide diversity of input data, such as number of classes, amount of labelled data, and texture patterns. A specifically designed interactive deep learning GPU training platform for image classification (Nvidia Digit) was employed in the experiments. It has shown efficiency in both training and testing. The results have shown that increasing the number of epochs leads to a higher accuracy rate, as expected. However, the convergence state is highly related to datasets. For the batch size evaluation, it has shown that a larger batch size slightly decreases the classification accuracy compared to a small batch size. For example, selecting the value 32 as the batch size on the RSCCN dataset achieves the accuracy rate of 90.34 % at the 11th epoch while decreasing the epoch value to one makes the accuracy rate drop to 74%. On the other extreme, setting an increased value of batch size to 200 decreases the accuracy rate at the 11th epoch is 86.5%, and 63% when using one epoch only. On the other hand, selecting the kernel size is loosely related to data set. From a practical point of view, the filter size 20 produces 70.4286%. The last performed image size experiment shows a dependency in the accuracy improvement. However, an expensive performance gain had been noticed. The represented conclusion opens the opportunities toward a better classification performance in various applications such as planetary remote sensing.

Keywords: CNNs, hyperparamters, remote sensing, land cover, land use

Procedia PDF Downloads 168

Authors: Paweł Konieczka, Lech Raczyński, Wojciech Wiślicki, Oleksandr Fedoruk, Konrad Klimaszewski, Przemysław Kopka, Wojciech Krzemień, Roman Shopa, Jakub Baran, Aurélien Coussat, Neha Chug, Catalina Curceanu, Eryk Czerwiński, Meysam Dadgar, Kamil Dulski, Aleksander Gajos, Beatrix C. Hiesmayr, Krzysztof Kacprzak, łukasz Kapłon, Grzegorz Korcyl, Tomasz Kozik, Deepak Kumar, Szymon Niedźwiecki, Dominik Panek, Szymon Parzych, Elena Pérez Del Río, Sushil Sharma, Shivani Shivani, Magdalena Skurzok, Ewa łucja Stępień, Faranak Tayefi, Paweł Moskal

Abstract:

This paper develops the transformation of non-image data into 2-dimensional matrices, as a preparation stage for classification based on convolutional neural networks (CNNs). In positron emission tomography (PET) studies, CNN may be applied directly to the reconstructed distribution of radioactive tracers injected into the patient's body, as a pattern recognition tool. Nonetheless, much PET data still exists in non-image format and this fact opens a question on whether they can be used for training CNN. In this contribution, the main focus of this paper is the problem of processing vectors with a small number of features in comparison to the number of pixels in the output images. The proposed methodology was applied to the classification of PET coincidence events.

Keywords: convolutional neural network, kernel principal component analysis, medical imaging, positron emission tomography

Procedia PDF Downloads 143

Authors: Mark Harmon, Abdolghani Ebrahimi, Patrick Lucey, Diego Klabjan

Abstract:

In this paper, we predict the likelihood of a player making a shot in basketball from multiagent trajectories. Previous approaches to similar problems center on hand-crafting features to capture domain-specific knowledge. Although intuitive, recent work in deep learning has shown, this approach is prone to missing important predictive features. To circumvent this issue, we present a convolutional neural network (CNN) approach where we initially represent the multiagent behavior as an image. To encode the adversarial nature of basketball, we use a multichannel image which we then feed into a CNN. Additionally, to capture the temporal aspect of the trajectories, we use “fading.” We find that this approach is superior to a traditional FFN model. By using gradient ascent, we were able to discover what the CNN filters look for during training. Last, we find that a combined FFN+CNN is the best performing network with an error rate of 39%.

Keywords: basketball, computer vision, image processing, convolutional neural network

Procedia PDF Downloads 153

Authors: Camille Charlebois-Plante, Marie-Ève Bourassa, Gaelle Dumel, Meriem Sabir, Louis De Beaumont

Abstract:

To the extent of our knowledge, there has been little interest in the chronic effects of mild traumatic brain injury (mTBI) on cognition during normal aging. This is rather surprising considering the impacts on daily and social functioning. In addition, sustaining a mTBI during late adulthood may increase the effect of normal biological aging in individuals who consider themselves normal and healthy. The objective of this study was to characterize the persistent neuropsychological repercussions of mTBI sustained during late adulthood, on average 12 months prior to testing. To this end, 35 mTBI patients and 42 controls between the ages of 50 and 69 completed an exhaustive neuropsychological assessment lasting three hours. All mTBI patients were asymptomatic and all participants had a score ≥ 27 at the MoCA. The evaluation consisted of 20 standardized neuropsychological tests measuring memory, attention, executive and language functions, as well as information processing speed. Performance on tests of visual (Brief Visuospatial Memory Test Revised) and verbal memory (Rey Auditory Verbal Learning Test and WMS-IV Logical Memory subtest), lexical access (Boston Naming Test) and response inhibition (Stroop) revealed to be significantly lower in the mTBI group. These findings suggest that a mTBI sustained during late adulthood induces lasting effects on cognitive function. Episodic memory and executive functions seem to be particularly vulnerable to enduring mTBI effects.

Keywords: cognitive function, late adulthood, mild traumatic brain injury, neuropsychology

Procedia PDF Downloads 169

Authors: J. Fox, J. Randhawa, M. Chan, L. Campbell, A. Weakely, D. J. Harvey, S. Tomaszewski Farias

Abstract:

Early identification of individuals at risk for conversion to dementia provides an opportunity for preventative treatment. Many older adults (30-60%) report specific subjective cognitive decline (SCD); however, previous research is inconsistent in terms of what types of complaints predict future cognitive decline. The purpose of this study is to identify which specific complaints from the Everyday Cognition Scales (ECog) scales, a measure of self-reported concerns for everyday abilities across six cognitive domains, are associated with: 1) conversion from a clinical diagnosis of normal to either MCI or dementia (categorical variable) and 2) progressive cognitive decline in memory and executive function (continuous variables). 415 cognitively normal older adults were monitored annually for an average of 5 years. Cox proportional hazards models were used to assess associations between self-reported ECog items and progression to impairment (MCI or dementia). A total of 114 individuals progressed to impairment; the mean time to progression was 4.9 years (SD=3.4 years, range=0.8-13.8). Follow-up models were run controlling for depression. A subset of individuals (n=352) underwent repeat cognitive assessments for an average of 5.3 years. For those individuals, mixed effects models with random intercepts and slopes were used to assess associations between ECog items and change in neuropsychological measures of episodic memory or executive function. Prior to controlling for depression, subjective concerns on five of the eight Everyday Memory items, three of the nine Everyday Language items, one of the seven Everyday Visuospatial items, two of the five Everyday Planning items, and one of the six Everyday Organization items were associated with subsequent diagnostic conversion (HR=1.25 to 1.59, p=0.003 to 0.03). However, after controlling for depression, only two specific complaints of remembering appointments, meetings, and engagements and understanding spoken directions and instructions were associated with subsequent diagnostic conversion. Episodic memory in individuals reporting no concern on ECog items did not significantly change over time (p>0.4). More complaints on seven of the eight Everyday Memory items, three of the nine Everyday Language items, and three of the seven Everyday Visuospatial items were associated with a decline in episodic memory (Interaction estimate=-0.055 to 0.001, p=0.003 to 0.04). Executive function in those reporting no concern on ECog items declined slightly (p <0.001 to 0.06). More complaints on three of the eight Everyday Memory items and three of the nine Everyday Language items were associated with a decline in executive function (Interaction estimate=-0.021 to -0.012, p=0.002 to 0.04). These findings suggest that specific complaints across several cognitive domains are associated with diagnostic conversion. Specific complaints in the domains of Everyday Memory and Language are associated with a decline in both episodic memory and executive function. Increased monitoring and treatment of individuals with these specific SCD may be warranted.

Keywords: alzheimer’s disease, dementia, memory complaints, mild cognitive impairment, risk factors, subjective cognitive decline

Procedia PDF Downloads 80

Authors: Wei Zhang

Abstract:

With the rapid development of deep learning, neural network and deep learning algorithms play a significant role in various practical applications. Due to the high accuracy and good performance, Convolutional Neural Networks (CNNs) especially have become a research hot spot in the past few years. However, the size of the networks becomes increasingly large scale due to the demands of the practical applications, which poses a significant challenge to construct a high-performance implementation of deep learning neural networks. Meanwhile, many of these application scenarios also have strict requirements on the performance and low-power consumption of hardware devices. Therefore, it is particularly critical to choose a moderate computing platform for hardware acceleration of CNNs. This article aimed to survey the recent advance in Field Programmable Gate Array (FPGA)-based acceleration of CNNs. Various designs and implementations of the accelerator based on FPGA under different devices and network models are overviewed, and the versions of Graphic Processing Units (GPUs), Application Specific Integrated Circuits (ASICs) and Digital Signal Processors (DSPs) are compared to present our own critical analysis and comments. Finally, we give a discussion on different perspectives of these acceleration and optimization methods on FPGA platforms to further explore the opportunities and challenges for future research. More helpfully, we give a prospect for future development of the FPGA-based accelerator.

Keywords: deep learning, field programmable gate array, FPGA, hardware accelerator, convolutional neural networks, CNN

Procedia PDF Downloads 128

Authors: Gazal Khan

Abstract:

This paper investigates the phenomenon of (re) creating and (re) posting of what is popularly known as "gaja-gamini walk" on instagram as a form of digital feminism, examining how these reels (short videos) make meaning in digital spaces. The study analyzes xyz “gaja- gamini walk” reels created by Indian influencers and instagram users, employing qualitative textual analysis, close readings, and digital ethnography to analyze the interplay between media, memory and digital spaces. The research highlights how “gaja-gamini walk” reels, characterized by an assertive presentation, redefines female body aesthetics, re (orients) sexual gaze to provide layered, interwoven and contested narratives. These reels facilitate a unique form of engagement by allowing users to re-share and participate in feminist discourse and allowing reels to function as sites of memory. The paper also discusses the social dynamics of these reels, their intertextuality with cultural narratives, and the limitations of the format for sustained feminist action. Through this analysis, the paper contributes to understanding the role of digital memory in contemporary feminist movements in context of Indian feminism.

Keywords: instagram, gaja-gamni walk, female gaze, digital feminism

Procedia PDF Downloads 33

Authors: Osman Adiguzel

Abstract:

Shape memory effect is a peculiar property exhibited by certain alloy systems and based on martensitic transformation, and shape memory properties are closely related to the microstructures of the material. Shape memory effect is linked with martensitic transformation, which is a solid state phase transformation and occurs with the cooperative movement of atoms by means of lattice invariant shears on cooling from high-temperature parent phase. Lattice twinning and detwinning can be considered as elementary processes activated during the transformation. Thermally induced martensite occurs as martensite variants, in self-accommodating manner and consists of lattice twins. Also, this martensite is called the twinned martensite or multivariant martensite. Deformation of shape memory alloys in martensitic state proceeds through a martensite variant reorientation. The martensite variants turn into the reoriented single variants with deformation, and the reorientation process has great importance for the shape memory behavior. Copper based alloys exhibit this property in metastable β- phase region, which has DO3 –type ordered lattice in ternary case at high temperature, and these structures martensiticaly turn into the layered complex structures with lattice twinning mechanism, on cooling from high temperature parent phase region. The twinning occurs as martensite variants with lattice invariant shears in two opposite directions, <110 > -type directions on the {110}- type plane of austenite matrix. Lattice invariant shear is not uniform in copper based ternary alloys and gives rise to the formation of unusual layered structures, like 3R, 9R, or 18R depending on the stacking sequences on the close-packed planes of the ordered lattice. The unit cell and periodicity are completed through 18 atomic layers in case of 18R-structure. On the other hand, the deformed material recovers the original shape on heating above the austenite finish temperature. Meanwhile, the material returns to the twinned martensite structures (thermally induced martensite structure) in one way (irreversible) shape memory effect on cooling below the martensite finish temperature, whereas the material returns to the detwinned martensite structure (deformed martensite) in two-way (reversible) shape memory effect. Shortly one can say that the microstructural mechanisms, responsible for the shape memory effect are the twinning and detwinning processes as well as martensitic transformation. In the present contribution, x-ray diffraction, transmission electron microscopy (TEM) and differential scanning calorimetry (DSC) studies were carried out on two copper-based ternary alloys, CuZnAl, and CuAlMn.

Keywords: shape memory effect, martensitic transformation, twinning and detwinning, layered structures

Procedia PDF Downloads 428

Authors: Dorottya Horváth, Tímea Harmath-Tánczos

Abstract:

Attention deficit hyperactivity disorder (ADHD) and learning disabilities are common neurocognitive disorders that can have a significant impact on a child's academic performance. ADHD is characterized by inattention, hyperactivity, and impulsivity, while learning disabilities are characterized by difficulty with specific academic skills, such as reading, writing, or math. The aim of this study was to investigate the working memory, executive, and attention functions of neurotypical children and children with ADHD and learning disabilities in order to fill the gaps in the Hungarian mean test scores of these cognitive functions in children with neurocognitive disorders. Another aim was to specify the neuropsychological differential diagnostic toolkit in terms of the relationships and peculiarities between these cognitive functions. The research question addressed in this study was: How do the working memory, executive, and attention functions of neurotypical children compare to those of children with ADHD and learning disabilities? A self-administered test battery was used as a research tool. Working memory was measured with the Non-Word Repetition Test, the Listening Span Test, the Digit Span Test, and the Reverse Digit Span Test; executive function with the Letter Fluency, Semantic Fluency, and Verb Fluency Tests; and attentional concentration with the d2-R Test. The data for this study was collected from 115 children aged 9-14 years. The children were divided into three groups: neurotypical children (n = 44), children with ADHD without learning disabilities (n = 23), and children with ADHD with learning disabilities (n = 48). The data was analyzed using a variety of statistical methods, including t-tests, ANOVAs, and correlational analyses. The results showed that the performance of children with neurocognitive involvement in working memory, executive functions, and attention was significantly lower than the performance of neurotypical children. However, the results of children with ADHD and ADHD with learning disabilities did not show a significant difference. The findings of this study are important because they provide new insights into the cognitive profiles of children with ADHD and learning disabilities and suggest that working memory, executive functions, and attention are all impaired in children with neurocognitive involvement, regardless of whether they have ADHD or learning disabilities. This information can be used to develop more effective diagnostic and treatment strategies for these disorders.

Keywords: ADHD, attention functions, executive functions, learning disabilities, working memory

Procedia PDF Downloads 94

Authors: Ousmane Youme, Jean Marie Dembele, Eugene Ezin, Christophe Cambier

Abstract:

In recent years, the convolutional neural networks (CNN) architectures designed by evolution algorithms have proven to be competitive with handcrafted architectures designed by experts. However, these algorithms need a lot of computational power, which is beyond the capabilities of most researchers and engineers. To overcome this problem, we propose an evolution architecture under length constraints. It consists of two algorithms: a search length strategy to find an optimal space and a search architecture strategy based on a genetic algorithm to find the best individual in the optimal space. Our algorithms drastically reduce resource costs and also keep good performance. On the Cifar-10 dataset, our framework presents outstanding performance with an error rate of 5.12% and only 4.6 GPU a day to converge to the optimal individual -22 GPU a day less than the lowest cost automatic evolutionary algorithm in the peer competition.

Keywords: CNN architecture, genetic algorithm, evolution algorithm, length constraints

Procedia PDF Downloads 128

Authors: Tian Lan, Yixiang Wang, Wenxin Tai, Yilan Lyu, Zufeng Wu

Abstract:

The monaural speech enhancement task in the time-frequencydomain has a myriad of approaches, with the stacked con-volutional neural network (CNN) demonstrating superiorability in feature extraction and selection. However, usingstacked single convolutions method limits feature represen-tation capability and generalization ability. In order to solvethe aforementioned problem, we propose an attention-basedadaptive convolutional network that integrates the multi-scale convolutional operations into a operation-specific blockvia input dependent attention to adapt to complex auditoryscenes. In addition, we introduce a two-stage progressivelearning method to enlarge the receptive field without a dra-matic increase in computation burden. We conduct a series ofexperiments based on the TIMIT corpus, and the experimen-tal results prove that our proposed model is better than thestate-of-art models on all metrics.

Keywords: speech enhancement, adaptive convolu-tion, progressive learning, time-frequency domain

Procedia PDF Downloads 122

Authors: Kavita Goel, Donald Winchester

Abstract:

In a world of information overload and work complexities, academics often struggle to create an online instructional environment enabling efficient and effective student learning. Research has established that students’ learning styles are different, some learn faster when taught using audio and visual methods. Attributes like prior knowledge and mental effort affect their learning. ‘Cognitive load theory’, opines learners have limited processing capacity. Cognitive load depends on the learner’s prior knowledge, the complexity of content and tasks, and instructional environment. Hence, the proper allocation of cognitive resources is critical for students’ learning. Consequently, a lecturer needs to understand the limits and strengths of the human learning processes, various learning styles of students, and accommodate these requirements while designing online assessments. As acknowledged in the cognitive load theory literature, visual and auditory explanations of worked examples potentially lead to a reduction of cognitive load (effort) and increased facilitation of learning when compared to conventional sequential text problem solving. This will help learner to utilize both subcomponents of their working memory. Instructional design changes were introduced at the case site for the delivery of the postgraduate business subjects. To make effective use of auditory and visual modalities, video recorded lectures, and key concept webinars were delivered to students. Videos were prepared to free up student limited working memory from irrelevant mental effort as all elements in a visual screening can be viewed simultaneously, processed quickly, and facilitates greater psychological processing efficiency. Most case study students in the postgraduate programs are adults, working full-time at higher management levels, and studying part-time. Their learning style and needs are different from other tertiary students. The purpose of the audio and visual interventions was to lower the students cognitive load and provide an online environment supportive to their efficient learning. These changes were expected to impact the student’s learning experience, their academic performance and retention favourably. This paper posits that these changes to instruction design facilitates students to integrate new knowledge into their long-term memory. A mixed methods case study methodology was used in this investigation. Primary data were collected from interviews and survey(s) of students and academics. Secondary data were collected from the organisation’s databases and reports. Some evidence was found that the academic performance of students does improve when new instructional design changes are introduced although not statistically significant. However, the overall grade distribution of student’s academic performance has changed and skewed higher which shows deeper understanding of the content. It was identified from feedback received from students that recorded webinars served as better learning aids than material with text alone, especially with more complex content. The recorded webinars on the subject content and assessments provides flexibility to students to access this material any time from repositories, many times, and this enhances students learning style. Visual and audio information enters student’s working memory more effectively. Also as each assessment included the application of the concepts, conceptual knowledge interacted with the pre-existing schema in the long-term memory and lowered student’s cognitive load.

Keywords: cognitive load theory, learning style, instructional environment, working memory

Procedia PDF Downloads 145

Authors: Xiaocong Liu, Huazhen Wang, Ting He, Xiaozheng Li, Weihan Zhang, Jian Chen

Abstract:

The utilization of electronic medical record (EMR) data to establish the disease diagnosis model has become an important research content of biomedical informatics. Deep learning can automatically extract features from the massive data, which brings about breakthroughs in the study of EMR data. The challenge is that deep learning lacks semantic knowledge, which leads to impracticability in medical science. This research proposes a method of incorporating lexical-semantic knowledge from abundant entities into a convolutional neural network (CNN) framework for pediatric disease diagnosis. Firstly, medical terms are vectorized into Lexical Semantic Vectors (LSV), which are concatenated with the embedded word vectors of word2vec to enrich the feature representation. Secondly, the semantic distribution of medical terms serves as Semantic Decision Guide (SDG) for the optimization of deep learning models. The study evaluate the performance of LSV-SDG-CNN model on four kinds of Chinese EMR datasets. Additionally, CNN, LSV-CNN, and SDG-CNN are designed as baseline models for comparison. The experimental results show that LSV-SDG-CNN model outperforms baseline models on four kinds of Chinese EMR datasets. The best configuration of the model yielded an F1 score of 86.20%. The results clearly demonstrate that CNN has been effectively guided and optimized by lexical-semantic knowledge, and LSV-SDG-CNN model improves the disease classification accuracy with a clear margin.

Keywords: convolutional neural network, electronic medical record, feature representation, lexical semantics, semantic decision

Procedia PDF Downloads 126

Authors: Jiangxia Han, Liang Xue, Keda Chen

Abstract:

Despite the significant progress over the last decades in reservoir simulation using numerical discretization, meshing is complex. Moreover, the high degree of freedom of the space-time flow field makes the solution process very time-consuming. Therefore, we present Physics-Informed Convolutional Neural Networks(PICNN) as a hybrid scientific theory and data method for reservoir modeling. Besides labeled data, the model is driven by the scientific theories of the underlying problem, such as governing equations, boundary conditions, and initial conditions. PICNN integrates governing equations and boundary conditions into the network architecture in the form of a customized convolution kernel. The loss function is composed of data matching, initial conditions, and other measurable prior knowledge. By customizing the convolution kernel and minimizing the loss function, the neural network parameters not only fit the data but also honor the governing equation. The PICNN provides a methodology to model and history-match flow and transport problems in porous media. Numerical results demonstrate that the proposed PICNN can provide an accurate physical solution from a limited dataset. We show how this method can be applied in the context of a forward simulation for continuous problems. Furthermore, several complex scenarios are tested, including the existence of data noise, different work schedules, and different good patterns.

Keywords: convolutional neural networks, deep learning, flow and transport in porous media, physics-informed neural networks, reservoir simulation

Procedia PDF Downloads 143

Authors: Ebipatei Victoria Tunyan, T. A. Cao, Cheol Young Ock

Abstract:

Detecting subjectively biased statements is a vital task. This is because this kind of bias, when present in the text or other forms of information dissemination media such as news, social media, scientific texts, and encyclopedias, can weaken trust in the information and stir conflicts amongst consumers. Subjective bias detection is also critical for many Natural Language Processing (NLP) tasks like sentiment analysis, opinion identification, and bias neutralization. Having a system that can adequately detect subjectivity in text will boost research in the above-mentioned areas significantly. It can also come in handy for platforms like Wikipedia, where the use of neutral language is of importance. The goal of this work is to identify the subjectively biased language in text on a sentence level. With machine learning, we can solve complex AI problems, making it a good fit for the problem of subjective bias detection. A key step in this approach is to train a classifier based on BERT (Bidirectional Encoder Representations from Transformers) as upstream model. BERT by itself can be used as a classifier; however, in this study, we use BERT as data preprocessor as well as an embedding generator for a Bi-LSTM (Bidirectional Long Short-Term Memory) network incorporated with attention mechanism. This approach produces a deeper and better classifier. We evaluate the effectiveness of our model using the Wiki Neutrality Corpus (WNC), which was compiled from Wikipedia edits that removed various biased instances from sentences as a benchmark dataset, with which we also compare our model to existing approaches. Experimental analysis indicates an improved performance, as our model achieved state-of-the-art accuracy in detecting subjective bias. This study focuses on the English language, but the model can be fine-tuned to accommodate other languages.

Keywords: subjective bias detection, machine learning, BERT–BiLSTM–Attention, text classification, natural language processing

Procedia PDF Downloads 130

Authors: Min Jindong, Wang Mingxia

Abstract:

With the widespread use of artificial intelligence in life, computer vision, especially deep convolutional neural network models, has developed rapidly. With the increase of the complexity of the real visual target detection task and the improvement of the recognition accuracy, the target detection network model is also very large. The huge deep neural network model is not conducive to deployment on edge devices with limited resources, and the timeliness of network model inference is poor. In this paper, knowledge distillation is used to compress the huge and complex deep neural network model, and the knowledge contained in the complex network model is comprehensively transferred to another lightweight network model. Different from traditional knowledge distillation methods, we propose a novel knowledge distillation that incorporates multi-faceted features, called M-KD. In this paper, when training and optimizing the deep neural network model for target detection, the knowledge of the soft target output of the teacher network in knowledge distillation, the relationship between the layers of the teacher network and the feature attention map of the hidden layer of the teacher network are transferred to the student network as all knowledge. in the model. At the same time, we also introduce an intermediate transition layer, that is, an intermediate guidance layer, between the teacher network and the student network to make up for the huge difference between the teacher network and the student network. Finally, this paper adds an exploration module to the traditional knowledge distillation teacher-student network model. The student network model not only inherits the knowledge of the teacher network but also explores some new knowledge and characteristics. Comprehensive experiments in this paper using different distillation parameter configurations across multiple datasets and convolutional neural network models demonstrate that our proposed new network model achieves substantial improvements in speed and accuracy performance.

Keywords: object detection, knowledge distillation, convolutional network, model compression

Procedia PDF Downloads 278