Search results for: variational auto-encoder

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 129

Authors: Alberto Hananel

Abstract:

The aim of this work is to modelize the occlusion of a person with temporomandibular disorders as an evolutionary equation and approach its solution by the construction and characterizing of discrete variational splines. To formulate the problem, certain boundary conditions have been considered. After showing the existence and the uniqueness of the solution of such a problem, a convergence result of a discrete variational evolutionary spline is shown. A stress analysis of the occlusion of a human jaw with temporomandibular disorders by finite elements is carried out in FreeFem++ in order to prove the validity of the presented method.

Keywords: approximation, evolutionary PDE, Finite Element Method, temporomandibular disorders, variational spline

Procedia PDF Downloads 351

Authors: Souvik Phadikar, Nidul Sinha, Rajdeep Ghosh

Abstract:

In brainwaves research using electroencephalogram (EEG) signals, finding the most relevant and effective feature set for identification of activities in the human brain is a big challenge till today because of the random nature of the signals. The feature extraction method is a key issue to solve this problem. Finding those features that prove to give distinctive pictures for different activities and similar for the same activities is very difficult, especially for the number of activities. The performance of a classifier accuracy depends on this quality of feature set. Further, more number of features result in high computational complexity and less number of features compromise with the lower performance. In this paper, a novel idea of the selection of optimal feature set using a heuristically optimized deep autoencoder is presented. Using various feature extraction methods, a vast number of features are extracted from the EEG signals and fed to the autoencoder deep neural network. The autoencoder encodes the input features into a small set of codes. To avoid the gradient vanish problem and normalization of the dataset, a meta-heuristic search algorithm is used to minimize the mean square error (MSE) between encoder input and decoder output. To reduce the feature set into a smaller one, 4 hidden layers are considered in the autoencoder network; hence it is called Heuristically Optimized Deep Autoencoder (HO-DAE). In this method, no features are rejected; all the features are combined into the response of responses of the hidden layer. The results reveal that higher accuracy can be achieved using optimal reduced features. The proposed HO-DAE is also compared with the regular autoencoder to test the performance of both. The performance of the proposed method is validated and compared with the other two methods recently reported in the literature, which reveals that the proposed method is far better than the other two methods in terms of classification accuracy.

Keywords: autoencoder, brainwave signal analysis, electroencephalogram, feature extraction, feature selection, optimization

Procedia PDF Downloads 95

Authors: Massimo Miccoli, Luca Marangoni, Alberto Aniello Scaringi, Alessandro Marceddu, Alessandro Amicone

Abstract:

The susceptibility of deep neural networks (DNNs) to adversarial manipulations is a recognized challenge within the computer vision domain. Adversarial examples, crafted by adding subtle yet malicious alterations to benign images, exploit this vulnerability. Various defense strategies have been proposed to safeguard DNNs against such attacks, stemming from diverse research hypotheses. Building upon prior work, our approach involves the utilization of autoencoder models. Autoencoders, a type of neural network, are trained to learn representations of training data and reconstruct inputs from these representations, typically minimizing reconstruction errors like mean squared error (MSE). Our autoencoder was trained on a dataset of benign examples; learning features specific to them. Consequently, when presented with significantly perturbed adversarial examples, the autoencoder exhibited high reconstruction errors. The architecture of the autoencoder was tailored to the dimensions of the images under evaluation. We considered various image sizes, constructing models differently for 256x256 and 512x512 images. Moreover, the choice of the computer vision model is crucial, as most adversarial attacks are designed with specific AI structures in mind. To mitigate this, we proposed a method to replace image-specific dimensions with a structure independent of both dimensions and neural network models, thereby enhancing robustness. Our multi-modal autoencoder reconstructs the spectral representation of images across the red-green-blue (RGB) color channels. To validate our approach, we conducted experiments using diverse datasets and subjected them to adversarial attacks using models such as ResNet50 and ViT_L_16 from the torch vision library. The autoencoder extracted features used in a classification model, resulting in an MSE (RGB) of 0.014, a classification accuracy of 97.33%, and a precision of 99%.

Keywords: adversarial attacks, malicious images detector, binary classifier, multimodal transformer autoencoder

Procedia PDF Downloads 52

Authors: C. Juntarasaid, T. Pulngern, S. Chucheepsakul

Abstract:

A formulation of postbuckling analysis of end supported rods under self-weight has been presented by the variational method. The variational formulation involving the strain energy due to bending and the potential energy of the self-weight, are expressed in terms of the intrinsic coordinates. The variational formulation is accomplished by introducing the Lagrange multiplier technique to impose the boundary conditions. The finite element method is used to derive a system of nonlinear equations resulting from the stationary of the total potential energy and then Newton-Raphson iterative procedure is applied to solve this system of equations. The numerical results demonstrate the postbluckled configurations of end supported rods under self-weight. This finite element method based on variational formulation expressed in term of intrinsic coordinate is highly recommended for postbuckling analysis of end-supported rods under self-weight.

Keywords: postbuckling, finite element method, variational method, intrinsic coordinate

Procedia PDF Downloads 130

Authors: M. O. Olayiwola

Abstract:

Numerical solutions of the generalized Burger-Fisher are obtained using a Modified Variational Iteration Method (MVIM) with minimal computational efforts. The computed results with this technique have been compared with other results. The present method is seen to be a very reliable alternative method to some existing techniques for such nonlinear problems.

Keywords: burger-fisher, modified variational iteration method, lagrange multiplier, Taylor’s series, partial differential equation

Procedia PDF Downloads 410

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 43

Authors: Athmane Bakhta, Sebastien Leclaire, David Vidal, Francois Bertrand, Mohamed Cheriet

Abstract:

This work presents a new three-dimensional variational model proposed for the simulation of ink seepage into paper sheets at the fiber level. The model, inspired by the Hising model, takes into account a finite volume of ink and describes the system state through gravity, cohesion, and adhesion force interactions. At the mesoscopic scale, the paper substrate is modeled using a discretized fiber structure generated using a numerical deposition procedure. A modified Monte Carlo method is introduced for the simulation of the ink dynamics. Besides, a multiphase lattice Boltzmann method is suggested to fine-tune the mesoscopic variational model parameters, and it is shown that the ink seepage behaviors predicted by the proposed model can resemble those predicted by a method relying on first principles.

Keywords: fibrous media, lattice Boltzmann, modelling and simulation, Monte Carlo, variational model

Procedia PDF Downloads 130

Authors: Azeddine El-Hassouny, Chandrashekhar Meshram, Geraldin Nanfack

Abstract:

In recent years, learning from data with noisy labels (Label Noise) has been a major concern in supervised learning. This problem has become even more worrying in Deep Learning, where the generalization capabilities have been questioned lately. Indeed, deep learning requires a large amount of data that is generally collected by search engines, which frequently return data with unreliable labels. In this paper, we investigate the Label Noise in Deep Learning using variational inference. Our contributions are : (1) exploiting Label Noise concept where the true labels are learnt using reparameterization variational inference, while observed labels are learnt discriminatively. (2) the noise transition matrix is learnt during the training without any particular process, neither heuristic nor preliminary phases. The theoretical results shows how true label distribution can be learned by variational inference in any discriminate neural network, and the effectiveness of our approach is proved in several target datasets, such as MNIST and CIFAR32.

Keywords: label noise, deep learning, discrete latent variable, variational inference, MNIST, CIFAR32

Procedia PDF Downloads 99

Authors: Kholod Mohammad Abualnaja

Abstract:

A new algorithm is presented to find some new iterative methods for solving nonlinear equations F(x)=0 by using the variational iteration method. The efficiency of the considered method is illustrated by example. The results show that the proposed iteration technique, without linearization or small perturbation, is very effective and convenient.

Keywords: variational iteration method, nonlinear equations, Lagrange multiplier, algorithms

Procedia PDF Downloads 519

Authors: Olayiwola Moruf Oyedunsi

Abstract:

This paper discusses the Modified Variational Iteration Method (MVIM) for the solution of nonlinear Burgers’ equation arising in longitudinal dispersion phenomena in fluid flow through porous media. The method is an elegant combination of Taylor’s series and the variational iteration method (VIM). Using Maple 18 for implementation, it is observed that the procedure provides rapidly convergent approximation with less computational efforts. The result shows that the concentration C(x,t) of the contaminated water decreases as distance x increases for the given time t.

Keywords: modified variational iteration method, Burger’s equation, porous media, partial differential equation

Procedia PDF Downloads 294

Authors: Elham Bagheri, Yalda Mohsenzadeh

Abstract:

Image memorability refers to the phenomenon where certain images are more likely to be remembered by humans than others. It is a quantifiable and intrinsic attribute of an image. Understanding how visual perception and memory interact is important in both cognitive science and artificial intelligence. It reveals the complex processes that support human cognition and helps to improve machine learning algorithms by mimicking the brain's efficient data processing and storage mechanisms. To explore the computational underpinnings of image memorability, this study examines the relationship between an image's reconstruction error, distinctiveness in latent space, and its memorability score. A trained autoencoder is used to replicate human-like memorability assessment inspired by the visual memory game employed in memorability estimations. This study leverages a VGG-based autoencoder that is pre-trained on the vast ImageNet dataset, enabling it to recognize patterns and features that are common to a wide and diverse range of images. An empirical analysis is conducted using the MemCat dataset, which includes 10,000 images from five broad categories: animals, sports, food, landscapes, and vehicles, along with their corresponding memorability scores. The memorability score assigned to each image represents the probability of that image being remembered by participants after a single exposure. The autoencoder is finetuned for one epoch with a batch size of one, attempting to create a scenario similar to human memorability experiments where memorability is quantified by the likelihood of an image being remembered after being seen only once. The reconstruction error, which is quantified as the difference between the original and reconstructed images, serves as a measure of how well the autoencoder has learned to represent the data. The reconstruction error of each image, the error reduction, and its distinctiveness in latent space are calculated and correlated with the memorability score. Distinctiveness is measured as the Euclidean distance between each image's latent representation and its nearest neighbor within the autoencoder's latent space. Different structural and perceptual loss functions are considered to quantify the reconstruction error. The results indicate that there is a strong correlation between the reconstruction error and the distinctiveness of images and their memorability scores. This suggests that images with more unique distinct features that challenge the autoencoder's compressive capacities are inherently more memorable. There is also a negative correlation between the reduction in reconstruction error compared to the autoencoder pre-trained on ImageNet, which suggests that highly memorable images are harder to reconstruct, probably due to having features that are more difficult to learn by the autoencoder. These insights suggest a new pathway for evaluating image memorability, which could potentially impact industries reliant on visual content and mark a step forward in merging the fields of artificial intelligence and cognitive science. The current research opens avenues for utilizing neural representations as instruments for understanding and predicting visual memory.

Keywords: autoencoder, computational vision, image memorability, image reconstruction, memory retention, reconstruction error, visual perception

Procedia PDF Downloads 54

Authors: Jiangyong Liu, Xiangxiang Xu, Bote Luo, Xiaoxue Luo, Jiang Zhu, Lingzhi Yi

Abstract:

To address the problems of non-linearity and high randomness of the original power load sequence causing the degradation of power load forecasting accuracy, a short-term load forecasting method is proposed. The method is based on the Least Square Support Vector Machine optimized by an Improved Sparrow Search Algorithm combined with the Variational Mode Decomposition proposed in this paper. The application of the variational mode decomposition technique decomposes the raw power load data into a series of Intrinsic Mode Functions components, which can reduce the complexity and instability of the raw data while overcoming modal confounding; the proposed improved sparrow search algorithm can solve the problem of difficult selection of learning parameters in the least Square Support Vector Machine. Finally, through comparison experiments, the results show that the method can effectively improve prediction accuracy.

Keywords: load forecasting, variational mode decomposition, improved sparrow search algorithm, least square support vector machine

Procedia PDF Downloads 78

Authors: Wanhyun Cho, Soonja Kang, Sangkyoon Kim, Soonyoung Park

Abstract:

In this paper, we present the human action recognition method using the variational Bayesian HMM with the Dirichlet process mixture (DPM) of the Gaussian-Wishart emission model (GWEM). First, we define the Bayesian HMM based on the Dirichlet process, which allows an infinite number of Gaussian-Wishart components to support continuous emission observations. Second, we have considered an efficient variational Bayesian inference method that can be applied to drive the posterior distribution of hidden variables and model parameters for the proposed model based on training data. And then we have derived the predictive distribution that may be used to classify new action. Third, the paper proposes a process of extracting appropriate spatial-temporal feature vectors that can be used to recognize a wide range of human behaviors from input video image. Finally, we have conducted experiments that can evaluate the performance of the proposed method. The experimental results show that the method presented is more efficient with human action recognition than existing methods.

Keywords: human action recognition, Bayesian HMM, Dirichlet process mixture model, Gaussian-Wishart emission model, Variational Bayesian inference, prior distribution and approximate posterior distribution, KTH dataset

Procedia PDF Downloads 325

Authors: A. W. Gbolagade, M. O. Olayiwola, K. O. Kareem

Abstract:

In this paper, further to our result on recent paper on the solution of nonlinear advection equations, we present further results on the nonlinear nonhomogeneous advection equations using a modified variational iteration method.

Keywords: lagrange multiplier, non-homogeneous equations, advection equations, mathematics

Procedia PDF Downloads 274

Authors: Victor Breux, Jérôme Boutet, Alain Goret, Viviane Cattin

Abstract:

Early detection of anomalies in data centers is important to reduce downtimes and the costs of periodic maintenance. However, there is little research on this topic and even fewer on the fusion of sensor data for the detection of abnormal events. The goal of this paper is to propose a method for anomaly detection in data centers by combining sensor data (temperature, humidity, power) and deep learning models. The model described in the paper uses one autoencoder per sensor to reconstruct the inputs. The auto-encoders contain Long-Short Term Memory (LSTM) layers and are trained using the normal samples of the relevant sensors selected by correlation analysis. The difference signal between the input and its reconstruction is then used to classify the samples using feature extraction and a random forest classifier. The data measured by the sensors of a data center between January 2019 and May 2020 are used to train the model, while the data between June 2020 and May 2021 are used to assess it. Performances of the model are assessed a posteriori through F1-score by comparing detected anomalies with the data center’s history. The proposed model outperforms the state-of-the-art reconstruction method, which uses only one autoencoder taking multivariate sequences and detects an anomaly with a threshold on the reconstruction error, with an F1-score of 83.60% compared to 24.16%.

Keywords: anomaly detection, autoencoder, data centers, deep learning

Procedia PDF Downloads 165

Authors: Desmond Adair, Kairat Ismailov, Martin Jaeger

Abstract:

Modelling of Timoshenko beams on elastic foundations has been widely used in the analysis of buildings, geotechnical problems, and, railway and aerospace structures. For the elastic foundation, the most widely used models are one-parameter mechanical models or two-parameter models to include continuity and cohesion of typical foundations, with the two-parameter usually considered the better of the two. Knowledge of free vibration characteristics of beams on an elastic foundation is considered necessary for optimal design solutions in many engineering applications, and in this work, the efficient and accurate variational iteration method is developed and used to calculate natural frequencies of a Timoshenko beam on a two-parameter foundation. The variational iteration method is a technique capable of dealing with some linear and non-linear problems in an easy and efficient way. The calculations are compared with those using a finite-element method and other analytical solutions, and it is shown that the results are accurate and are obtained efficiently. It is found that the effect of the presence of the two-parameter foundation is to increase the beam’s natural frequencies and this is thought to be because of the shear-layer stiffness, which has an effect on the elastic stiffness. By setting the two-parameter model’s stiffness parameter to zero, it is possible to obtain a one-parameter foundation model, and so, comparison between the two foundation models is also made.

Keywords: Timoshenko beam, variational iteration method, two-parameter elastic foundation model

Procedia PDF Downloads 173

Authors: Harendra Singh, Rajesh Pandey

Abstract:

The paper presents a numerical method based on operational matrix of integration and Ryleigh method for the solution of a class of non-linear fractional variational problems (NLFVPs). Chebyshev first kind polynomials are used for the construction of operational matrix. Using operational matrix and Ryleigh method the NLFVP is converted into a system of non-linear algebraic equations, and solving these equations we obtained approximate solution for NLFVPs. Convergence analysis of the proposed method is provided. Numerical experiment is done to show the applicability of the proposed numerical method. The obtained numerical results are compared with exact solution and solution obtained from Chebyshev third kind. Further the results are shown graphically for different fractional order involved in the problems.

Keywords: non-linear fractional variational problems, Rayleigh-Ritz method, convergence analysis, error analysis

Procedia PDF Downloads 272

Authors: Shriya Shukla, Lachin Fernando

Abstract:

Pneumonia, a leading cause of mortality in young children globally, presents significant diagnostic challenges, particularly in resource-limited settings. This study presents an approach to diagnosing pediatric pneumonia using Chest X-Ray (CXR) images, employing a lightweight MobileNetV2 model enhanced with synthetic data augmentation. Addressing the challenge of dataset scarcity and imbalance, the study used a Variational Autoencoder-Generative Adversarial Network (VAE-GAN) to generate synthetic CXR images, improving the representation of normal cases in the pediatric dataset. This approach not only addresses the issues of data imbalance and scarcity prevalent in medical imaging but also provides a more accessible and reliable diagnostic tool for early pneumonia detection. The augmented data improved the model’s accuracy and generalization, achieving an overall accuracy of 95% in pneumonia detection. These findings highlight the efficacy of the MobileNetV2 model, offering a computationally efficient yet robust solution well-suited for resource-constrained environments such as mobile health applications. This study demonstrates the potential of synthetic data augmentation in enhancing medical image analysis for critical conditions like pediatric pneumonia.

Keywords: pneumonia, MobileNetV2, image classification, GAN, VAE, deep learning

Procedia PDF Downloads 51

Authors: Pushpendra Kumar, Sanjeev Kumar, R. Balasubramanian

Abstract:

Estimation of optical flow from a sequence of images using variational methods is one of the most successful approach. Discontinuity between different motions is one of the challenging problem in flow estimation. In this paper, we design a new anisotropic diffusion operator, which is able to provide smooth flow over a region and efficiently preserve discontinuity in optical flow. This operator is designed on the basis of intensity differences of the pixels and isotropic operator using exponential function. The combination of these are used to control the propagation of flow. Experimental results on the different datasets verify the robustness and accuracy of the algorithm and also validate the effect of anisotropic operator in the discontinuity preserving.

Keywords: optical flow, variational methods, computer vision, anisotropic operator

Procedia PDF Downloads 846

Authors: Surendra Mund

Abstract:

At the beginning, the physical entity force was defined mathematically by Sir Isaac Newton in his Principia Mathematica as F ⃗=(dp ⃗)/dt in form of his second law of motion. Newton also defines his Universal law of Gravitational force exist in same outstanding book, but at the end of 20th century and beginning of 21st century, we have tried a lot to specify and unify four or five Fundamental forces or Interaction exist in universe, but we failed every time. Usually, Gravity creates problems in this unification every single time, but in my previous papers and presentations, I defined and derived Field and force equations for Gravitational like Interactions for each and every kind of central systems. This force is named as Variational Force by me, and this force is generated by variation in the scalar field density around the body. In this particular paper, at first, I am specifying which type of Interactions are Fundamental in Universal sense (or in all type of central systems or bodies predicted by my N-time Inflationary Model of Universe) and then unify them in Universal framework (defined and derived by me as Universal Mechanics in a separate paper) as well. This will also be valid in Universal dynamical sense which includes inflations and deflations of universe, central system relativity, Universal relativity, ϕ-ψ transformation and transformation of spin, physical perception principle, Generalized Fundamental Dynamical Law and many other important Generalized Principles of Generalized Quantum Mechanics (GQM) and Central System Theory (CST). So, In this article, at first, I am Generalizing some Fundamental Principles, and then Unifying Variational Forces (General form of Gravitation like Interactions) and Flow Generated Force (General form of EM like Interactions), and then Unify all Fundamental Forces by specifying Weak and Strong Interactions in form of more basic terms - Variational, Flow Generated and Transformational Interactions.

Keywords: Central System Force, Disturbance Force, Flow Generated Forces, Generalized Nuclear Force, Generalized Weak Interactions, Generalized EM-Like Interactions, Imbalance Force, Spin Generated Forces, Transformation Generated Force, Unified Force, Universal Mechanics, Uniform And Non-Uniform Variational Interactions, Variational Interactions

Procedia PDF Downloads 29

Authors: Aadhya Kaul, Gautam Manocha, Preeti Nagrath

Abstract:

Our environment comprises of a wide variety of plants that are similar to each other and sometimes the similarity between the plants makes the identification process tedious thus increasing the workload of the botanist all over the world. Now all the botanists cannot be accessible all the time for such laborious plant identification; therefore, there is an urge for a quick classification model. Also, along with the identification of the plants, it is also necessary to classify the plant as healthy or not as for a good lifestyle, humans require good food and this food comes from healthy plants. A large number of techniques have been applied to classify the plants as healthy or diseased in order to provide the solution. This paper proposes one such method known as anomaly detection using autoencoders using a set of collections of leaves. In this method, an autoencoder model is built using Keras and then the reconstruction of the original images of the leaves is done and the threshold loss is found in order to classify the plant leaves as healthy or diseased. A dataset of plant leaves is considered to judge the reconstructed performance by convolutional autoencoders and the average accuracy obtained is 71.55% for the purpose.

Keywords: convolutional autoencoder, anomaly detection, web application, FLASK

Procedia PDF Downloads 139

Authors: O. Acan, Y. Keskin

Abstract:

In this study, we will carry out a comparative study between the reduced differential transform method, the adomian decomposition method, the variational iteration method and the homotopy analysis method. These methods are used in many fields of engineering. This is been achieved by handling a kind of 2-Dimensional and forth-order partial differential equations called the Kuramoto–Sivashinsky equations. Three numerical examples have also been carried out to validate and demonstrate efficiency of the four methods. Furthermost, it is shown that the reduced differential transform method has advantage over other methods. This method is very effective and simple and could be applied for nonlinear problems which used in engineering.

Keywords: reduced differential transform method, adomian decomposition method, variational iteration method, homotopy analysis method

Procedia PDF Downloads 410

Authors: Zhen Cheng, Xinyu Dai, Shujian Huang, Jiajun Chen

Abstract:

Recently, explanatory natural language inference has attracted much attention for the interpretability of logic relationship prediction, which is also known as explanation generation for Natural Language Inference (NLI). Existing explanation generators based on discriminative Encoder-Decoder architecture have achieved noticeable results. However, we find that these discriminative generators usually generate explanations with correct evidence but incorrect logic semantic. It is due to that logic information is implicitly encoded in the premise-hypothesis pairs and difficult to model. Actually, logic information identically exists between premise-hypothesis pair and explanation. And it is easy to extract logic information that is explicitly contained in the target explanation. Hence we assume that there exists a latent space of logic information while generating explanations. Specifically, we propose a generative model called Variational Explanation Generator (VariationalEG) with a latent variable to model this space. Training with the guide of explicit logic information in target explanations, latent variable in VariationalEG could capture the implicit logic information in premise-hypothesis pairs effectively. Additionally, to tackle the problem of posterior collapse while training VariaztionalEG, we propose a simple yet effective approach called Logic Supervision on the latent variable to force it to encode logic information. Experiments on explanation generation benchmark—explanation-Stanford Natural Language Inference (e-SNLI) demonstrate that the proposed VariationalEG achieves significant improvement compared to previous studies and yields a state-of-the-art result. Furthermore, we perform the analysis of generated explanations to demonstrate the effect of the latent variable.

Keywords: natural language inference, explanation generation, variational auto-encoder, generative model

Procedia PDF Downloads 126

Authors: Sandeep Kumar, Naveen Gupta

Abstract:

The propagation of the Non-Gaussian laser beam results in strong self-focusing as compare to the Gaussian laser beam, which helps to achieve a prerequisite of the plasma-based electron, Terahertz generation, and higher harmonic generations. The theoretical investigation on the evolution of non-Gaussian laser beam through the collisional plasma with ramped density has been presented. The non-uniform irradiance over the cross-section of the laser beam results in redistribution of the carriers that modifies the optical response of the plasma in such a way that the plasma behaves like a converging lens to the laser beam. The formulation is based on finding a semi-analytical solution of the nonlinear Schrodinger wave equation (NLSE) with the help of variational theory. It has been observed that the decentred parameter ‘q’ of laser and wavenumber of ripples of medium contribute to providing the required conditions for the improvement of self-focusing.

Keywords: non-Gaussian beam, collisional plasma, variational theory, self-focusing

Procedia PDF Downloads 172

Authors: Wanhyun Cho, Seongchae Seo, Soonja Kang

Abstract:

In this work, we propose a variational technique for image contrast enhancement which utilizes global and local information around each pixel. The energy functional is defined by a weighted linear combination of three terms which are called on a local, a global contrast term and dispersion term. The first one is a local contrast term that can lead to improve the contrast of an input image by increasing the grey-level differences between each pixel and its neighboring to utilize contextual information around each pixel. The second one is global contrast term, which can lead to enhance a contrast of image by minimizing the difference between its empirical distribution function and a cumulative distribution function to make the probability distribution of pixel values becoming a symmetric distribution about median. The third one is a dispersion term that controls the departure between new pixel value and pixel value of original image while preserving original image characteristics as well as possible. Second, we derive the Euler-Lagrange equation for true image that can achieve the minimum of a proposed functional by using the fundamental lemma for the calculus of variations. And, we considered the procedure that this equation can be solved by using a gradient decent method, which is one of the dynamic approximation techniques. Finally, by conducting various experiments, we can demonstrate that the proposed method can enhance the contrast of colour images better than existing techniques.

Keywords: color image, contrast enhancement technique, variational approach, Euler-Lagrang equation, dynamic approximation method, EME measure

Procedia PDF Downloads 429

Authors: El Hassene Osmani, Mounir Haddou, Naceurdine Bensalem

Abstract:

In this paper, we investigate optimal control problems governed by semilinear elliptic variational inequalities involving constraints on the state, and more precisely, the obstacle problem. We present a relaxed formulation for the problem using smoothing functions. Since we adopt a numerical point of view, we first relax the feasible domain of the problem, then using both mathematical programming methods and penalization methods, we get optimality conditions with smooth Lagrange multipliers. Some numerical experiments using IPOPT algorithm (Interior Point Optimizer) are presented to verify the efficiency of our approach.

Keywords: complementarity problem, IPOPT, Lagrange multipliers, mathematical programming, optimal control, smoothing methods, variationally inequalities

Procedia PDF Downloads 147

Authors: T. Bryan , V. Kepuska, I. Kostnaic

Abstract:

A method of data compression and denoising is introduced that is based on atomic decomposition of audio data using “basis vectors” that are learned from the audio data itself. The basis vectors are shown to have higher data compression and better signal-to-noise enhancement than the Gabor and gammatone “seed atoms” that were used to generate them. The basis vectors are the input weights of a Sparse AutoEncoder (SAE) that is trained using “envelope samples” of windowed segments of the audio data. The envelope samples are extracted from the audio data by performing atomic decomposition with Gabor or gammatone seed atoms. This process identifies segments of audio data that are locally coherent with the seed atoms. Envelope samples are extracted by identifying locally coherent audio data segments with Gabor or gammatone seed atoms, found by matching pursuit. The envelope samples are formed by taking the kronecker products of the atomic envelopes with the locally coherent data segments. Oracle signal-to-noise ratio (SNR) verses data compression curves are generated for the seed atoms as well as the basis vectors learned from Gabor and gammatone seed atoms. SNR data compression curves are generated for speech signals as well as early American music recordings. The basis vectors are shown to have higher denoising capability for data compression rates ranging from 90% to 99.84% for speech as well as music. Envelope samples are displayed as images by folding the time series into column vectors. This display method is used to compare of the output of the SAE with the envelope samples that produced them. The basis vectors are also displayed as images. Sparsity is shown to play an important role in producing the highest denoising basis vectors.

Keywords: sparse dictionary learning, autoencoder, sparse autoencoder, basis vectors, atomic decomposition, envelope sampling, envelope samples, Gabor, gammatone, matching pursuit

Procedia PDF Downloads 231

Authors: Rossella Arcucci, Luisa D'Amore, Simone Celestino, Giuseppe Scotti, Giuliano Laccetti

Abstract:

This work is the first dowel in a rather wide research activity in collaboration with Euro Mediterranean Center for Climate Changes, aimed at introducing scalable approaches in Ocean Circulation Models. We discuss designing and implementation of a parallel algorithm for solving the Variational Data Assimilation (DA) problem on Graphics Processing Units (GPUs). The algorithm is based on the fully scalable 3DVar DA model, previously proposed by the authors, which uses a Domain Decomposition approach (we refer to this model as the DD-DA model). We proceed with an incremental porting process consisting of 3 distinct stages: requirements and source code analysis, incremental development of CUDA kernels, testing and optimization. Experiments confirm the theoretic performance analysis based on the so-called scale up factor demonstrating that the DD-DA model can be suitably mapped on GPU architectures.

Keywords: data assimilation, GPU architectures, ocean models, parallel algorithm

Procedia PDF Downloads 387

Authors: Wanhyun Cho, Soonja Kang, Sanggoon Kim, Soonyoung Park

Abstract:

We present probabilistic multinomial Dirichlet classification model for multidimensional data and Gaussian process priors. Here, we have considered an efficient computational method that can be used to obtain the approximate posteriors for latent variables and parameters needed to define the multiclass Gaussian process classification model. We first investigated the process of inducing a posterior distribution for various parameters and latent function by using the variational Bayesian approximations and important sampling method, and next we derived a predictive distribution of latent function needed to classify new samples. The proposed model is applied to classify the synthetic multivariate dataset in order to verify the performance of our model. Experiment result shows that our model is more accurate than the other approximation methods.

Keywords: multinomial dirichlet classification model, Gaussian process priors, variational Bayesian approximation, importance sampling, approximate posterior distribution, marginal likelihood evidence

Procedia PDF Downloads 416