Deep Neural Networks (DNN)

Space exploration has advanced much further chief of all advancements made in rocketry. AI is
important in the rocket’s launch and landing in that they are able to improve on the existing systems. This
paper provides a critical discussion on the incorporation of AI technologies into these systems with reference
to effectiveness, safety, and mission accomplishment. Algorithms become an efficient way to assist decisions in
launch phases, to specify trajectories and to tweak auto-landing systems. Modern complex systems can be
analyzed by means of machine learning and neural networks, resulting in improved prediction of system
health and less maintenance failures. In addition, I believe that it is crucial that AI is used in controlling
reusability of rockets so as to remove human factor and at the same time bring in efficiency. The audit shows
that AI-powered algorithms have enhanced the launch accuracy up to 15% and minimized system failures up
to 10%. AI based autonomous landing systems have reported a 20% improvement in the accuracy of landing thereby reducing lifecycle risks and improving the reusability of rocket stages. It has been marked that integration of AI for diagnosing problems results in 25% overall system reliability therefore the study supports the statement. It has also enhanced the optimization of fuel, which is critical for mission sustainability, through a 12% improvement on fuel efficacies.

The security of modern communication networks cannot be adequately assured without intrusion detection systems (IDS). Pattern recognition, signature analysis, and rule violation detection were the primary goals of these systems. Recent advances in ML and DL approaches have shown promise as prospective replacements in a field of NID. Typical and anomalous patterns may be distinguished using these techniques. This paper uses the NSL-KDD benchmark data set to assess NIDS using many ML algorithms, like SVM, DT, LR, and RF classification. We evaluate the precision, recall, accuracy, and FPR of several ML techniques, such as SVM accuracy is 98.97%, Random Forests, and Decision Trees. The results demonstrate that, in comparison to conventional techniques, machine learning approaches greatly increase detection rates while reducing false alarms. And in this, a RF achieve a high accuracy which is 99.83%. The results of this investigation demonstrate that not only is it feasible to obtain a high detection rate of assaults, but also accurate

Representation learning is a transformative paradigm in machine learning, which is a ground breaking transformation from traditional feature engineering approaches to automatic discovery of features. In this comprehensive review, we look at the evolution of representation learning methods as applied to modern artificial intelligence systems, as well as their implications for both practitioners and researchers. With popular machine learning applications getting more complex, a need to automatically learn high level representations out of raw data is important for making the field go beyond manual feature engineering.

Semantic segmentation has major benefits in autonomous driving and robotics related applications, where scene understanding is a necessity. Most of the research on semantic segmentation is focused on increasing the accuracy of segmentation models with few research on real-time performance. The few work conducted in this direction does not also provide principled methods to evaluate the different design choices for segmentation. In this paper, we address this gap by presenting the first real-time semantic segmentation benchmarking framework 2. The framework is comprised of different network architectures for feature extraction such as VGG16, MobileNet, and ResNet-18. It is also comprised of multiple metaarchitectures for segmentation that define the decoding methodology. These include Skip architecture, UNet, and Dilation Frontend. Experimental results on cityscapes with a case study using MobileNet architecture and two meta-architectures are presented.

The field of explainable artificial intelligence (XAI) aims to explain the decisions of DNNs. Complete DNN explanations accurately reflect the inner workings of the DNN while interpretable explanations are easy for humans to understand. Developing methods for explaining the representations learned by DNNs that are both complete and interpretable is a grand challenge in the field of XAI. This thesis makes contributions to the field of XAI by proposing and evaluating novel methods for studying DNN representations. During forward propagation, each DNN layer non-linearly transforms the input space in some way that is useful for minimizing a loss function. To understand how DNNs represent their inputs, this work develops methods to examine each DNN layer’s activation space. The first article contributes an unsupervised framework for identifying and interpreting “tuning dimensions” in the activation space of DNN layers. The method consists of fitting a dimensionality reduction model to a ...

One indicator of a university's educational quality is the proportion of enrolled students who actually graduate within four years. This proportion is typically fewer than the number of students that enroll in a given year. A low graduation rate can have a negative impact on both the university's reputation and its accreditation because it indicates that fewer students are completing their degrees. Student activity, economic, and other issues all play a role in why some students are unable to complete their degrees on time. As a result, stakeholders need a model that can predict whether or not students will graduate on time as a means of evaluating and giving a basis for policy actions. This research proposes a model for converting textual data into an image format using a deep learning convolutional neural network (CNN), and then classifying the extracted features using a variety of machine learning classification algorithms like the decision tree, random forest, Naive Bayes, support vector machine (SVM), and k-nearest neighbor (K-NN). The classification model trained on feature extraction data had a 96.1% accuracy rate, while the classification model trained on the original data achieved a 71.2% accuracy rate.

The interpretability of deep neural networks (DNNs) has become a crucial focus within artificial intelligence and machine learning, particularly as these models are increasingly used in high-stakes applications such as healthcare, finance, and autonomous driving. This article explores the impact of architectural design choices on the interpretability of DNNs, emphasizing the importance of transparency, trust, and accountability in AI systems. By presenting case studies and experimental results, the article highlights how different architectural elements-such as layer types, network depth, connectivity patterns, and attention mechanisms-affect model interpretability and performance. The discussion is structured into three main sections: real-world applications, architectural trade-offs, and tools and techniques in practice. In healthcare, for example, interpretability techniques like heatmaps enhance diagnostic model transparency, aiding clinical decision-making and improving patient outcomes. In finance, methods such as LIME and SHAP provide clear explanations for credit scoring models, facilitating regulatory compliance and risk assessment. For autonomous driving, the article examines how interpretability ensures safety and reliability, fostering public trust. Through a comprehensive analysis of attention mechanisms and a comparison of convolutional versus recurrent layers, the article offers insights into balancing performance and interpretability. Additionally, it reviews practical tools like LIME and SHAP, demonstrating their effectiveness in enhancing model transparency. The findings underscore the necessity of tailored interpretability solutions to ensure the broader acceptance and effective utilization of DNNs in critical real-world scenarios.

The interpretability of deep neural networks (DNNs) is a critical focus in artificial intelligence (AI) and machine learning (ML), particularly as these models are increasingly deployed in high-stakes applications such as healthcare, finance, and autonomous systems. In the context of these technologies, interpretability refers to the extent to which a human can understand the cause of a decision made by a model. This article evaluates various methods for assessing the interpretability of DNNs, recognizing the significant challenges posed by their complex and opaque nature. The review encompasses both quantitative metrics and qualitative evaluations, aiming to identify effective strategies that enhance model transparency without compromising performance. The structure of the article includes an exploration of quantitative metrics, such as model complexity and computational requirements, and techniques like Local Interpretable Model-agnostic Explanations (LIME) and SHapley Additive exPlanations (SHAP). It also considers qualitative evaluations, emphasizing the role of human judgment through domain expert reviews and user studies. Additionally, the article addresses the necessity of standardized benchmarks and the importance of context-specific evaluation frameworks. Through an investigation of these approaches, the treatment provides an overview of current methods and proposes future directions for improving the interpretability of DNNs, thus enhancing trust, accountability, and transparency in AI systems.

The FOREX market assessment is a big challenge for investors and global risk managers. However, the present study uses daily multicurrency exchange rate returns data from 2007 to 2022 to estimate the learning returns performance of the proposed model to find a safe-haven currency for optimal investment strategy. The categorical returns are classified into good returns (GRs), bad returns (BRs) and no returns (NRs). Therefore, the present study needs to use a one-hot-encoding function to convert a categorical dataset into a numeric format with TensorFlow. The present study proposes a deep neural network-based multilayer perceptron (DNN-based MLP) with a backpropagation algorithm to estimate the learning returns performance of the proposed model to find a safe-haven currency for optimal investment strategy. The findings showed that currency exchange rate return 2 (CERR2) is relatively a safe-haven currency than currency exchange rate return 1 (CERR1) and currency exchange rate return 3 (CERR3). Moreover, the findings also showed that the proposed model gives optimal learning return performance. This study may assist FOREX investors to modify their investment strategies under shed light of findings of the study. In addition, the findings of the present study may also support global risk managers to revisit their hedging strategies.

Object classification problem is essential in many applications nowadays. Human can easily classify objects in unconstrained environments easily. Classical classification techniques were far away from human performance. Thus, researchers try to mimic the human visual system till they reached the deep neural networks. This chapter gives a review and analysis in the field of the deep convolutional neural network usage in object classification under constrained and unconstrained environment. The chapter gives a brief review on the classical techniques of object classification and the development of bioinspired computational models from neuroscience till the creation of deep neural networks. A review is given on the constrained environment issues: the hardware computing resources and memory, the object appearance and background, and the training and processing time. Datasets that are used to test the performance are analyzed according to the images environmental conditions, besides the dataset biasing is discussed.

Protease is a proteolytic enzyme that hydrolyzes the amino acid where the cleavage only occurs at specific sites of the amino acid substrate.  By discovering the nick site, the prediction on the function of proteases can be identified and enable humans to control the protein's hydrolysis by their corresponding protease. It is very contributed to controlling protein production especially viral protein. The experts may alter the production of viral protein by reducing the viral proteases to undergo proteolysis. With the rise of computational methods in this era, deep learning is becoming more famous and applied in every field of study, including the biological area. Conventional techniques such as mass spectrometry and two-dimensional gel electrophoresis are being replaced by computational methods due to time-consuming. Thus, this study improves the deep learning algorithm by proposing the Hybrid model of Random Forest + Deep Neural Network (Hybrid RF+DNN) to classify nick sites. ...

Recently, a generative variational autoencoder (VAE) has been proposed for speech enhancement to model speech statistics. However, this approach only uses clean speech in the training phase, making the estimation particularly sensitive to noise presence, especially in low signal-to-noise ratios (SNRs). To increase the robustness of the VAE, we propose to include noise information in the training phase by using a noise-aware encoder trained on noisy-clean speech pairs. We evaluate our approach on real recordings of different noisy environments and acoustic conditions using two different noise datasets. We show that our proposed noise-aware VAE outperforms the standard VAE in terms of overall distortion without increasing the number of model parameters. At the same time, we demonstrate that our model is capable of generalizing to unseen noise conditions better than a supervised feedforward deep neural network (DNN). Furthermore, we demonstrate the robustness of the model performance to a reduction of the noisy-clean speech training data size.

A cognitive agent performing in the real world needs to learn relevant concepts about its environment (e.g., objects, color, and shapes) and react accordingly. In addition to learning the concepts, it needs to learn relations between the concepts, in particular spatial relations between objects. In this paper, we propose three approaches that allow a cognitive agent to learn spatial relations. First, using an embodied model, the agent learns to reach toward an object based on simple instructions involving left-right relations. Since the level of realism and its complexity does not permit large-scale and diverse experiences in this approach, we devise as a second approach a simple visual dataset for geometric feature learning and show that recent reasoning models can learn directional relations in different frames of reference. Yet, embodied and simple simulation approaches together still do not provide sufficient experiences. To close this gap, we thirdly propose utilizing knowledge...

In this paper we propose and investigate a novel nonlinear unit, called Lp unit, for deep neural networks. The proposed Lp unit receives signals from several projections of a subset of units in the layer below and computes a normalized Lp norm. We notice two interesting interpretations of the Lp unit. First, the proposed unit can be understood as a generalization of a number of conventional pooling operators such as average, root-mean-square and max pooling widely used in, for instance, convolutional neural networks (CNN), HMAX models and neocognitrons. Furthermore, the Lp unit is, to a certain degree, similar to the recently proposed maxout unit [13] which achieved the state-of-the-art object recognition results on a number of benchmark datasets. Secondly, we provide a geometrical interpretation of the activation function based on which we argue that the Lp unit is more efficient at representing complex, nonlinear separating boundaries. Each Lp unit defines a superelliptic boundary, with its exact shape defined by the order p. We claim that this makes it possible to model arbitrarily shaped, curved boundaries more efficiently by combining a few Lp units of different orders. This insight justifies the need for learning different orders for each unit in the model. We empirically evaluate the proposed Lp units on a number of datasets and show that multilayer perceptrons (MLP) consisting of the Lp units achieve the state-of-the-art results on a number of benchmark datasets. Furthermore, we evaluate the proposed Lp unit on the recently proposed deep recurrent neural networks (RNN).

Due to the increasing deployment of vehicles in human societies and the necessity for smart traffic control, anomaly detection is among the various tasks widely employed in traffic monitoring. As the issue of urban traffic and their relative smart monitoring systems have gained popularity among researchers in recent years, there exist several studies in this regard. In most of these studies, classification is performed based on the behavior of drivers, where a set of default trajectories are used in order to learn the system and classify the related data. However, two understudied challenges are the lack of access to sufficient data to provide an efficient model, along with the lack of access to anomaly data that covers all possible abnormal trajectories. While the former challenge can be tackled through long-term data recording, the latter requires appropriate considerations. To this aim, we have utilized a combination of optimized convolutional neural network and fuzzy neural network classifiers, along with autoencoding neural networks. The final combination occurs at the decision level. First, the CNN-ANFIS classifier assigns the input trajectory to one of the predefined categories. Then, the trained autoencoder networks examine the result in order to find whether the trajectory is normal or abnormal. Obtaining 87.5% accuracy on QMUL and 99.5% on the T15 datasets confirms the superior performance of the proposed method.

Protease is a proteolytic enzyme that hydrolyzes the amino acid where the cleavage only occurs at specific sites of the amino acid substrate. By discovering the nick site, the prediction on the function of proteases can be identified and enable humans to control the protein’s hydrolysis by their corresponding protease. This is an important process to control as it can help to control protein replication especially viral proteins. With the rise of computational methods in this era, mainly through the successful application of deep learning in various domains, the application of this method in biological data can help to improve predictions to support the experimental methods. Conventional techniques such as mass spectrometry and two-dimensional gel electrophoresis can be supported by computational methods by preparing predictions. Thus reducing the cost of experiment and time taken to identify and predict the protein proteolysis site. This study improves the deep learning algorithm by proposing the Hybrid model of Random Forest + Deep Neural Network (Hybrid RF+DNN) to classify proteolysis or nick sites. The classification in
this study is compared with other machine learning algorithms such as Random Forest (RF), Support Vector Machine (SVM), and Deep Neural Network (DNN). The proposed method enhances the classification results in identifying the positive and negative nick sites. The RF is a feature selector that gathers the most important feature before entering the DNN classifier. This approach reduces the data dimensionality and speeds up the execution time of the training process. The performance of the models was measured by confusion matrix, specificity, sensitivity, etc. However, the proposed method is not the best performer among the mentioned classifiers as the classifiers have obtained 0.64, 0.65, and 0.58 for Datasets A, B, and C, respectively. The proposed method may become the best performer as the parameter tuning is done more precisely, even after the feature selection by the RF algorithm. Thus, the proposed method with the enhancement appears to be an alternative to the researcher discovering the limited proteolysis or nick site.

Currently, indoor home object recognition systems lack the degree of accuracy required for reliable automated operations. In this paper, a 3-Dimensional (3D) object recognition deep neural network system, capable of recognizing indoor objects from 3D images with a view to assisting indoor robotic devices in performing home tasks, is presented. Almost invariably such systems drastically increase the number of total parameters which must be trained leading to very timeconsuming training processes. Furthermore, the lack of suitably large annotated datasets for indoor objects adds to the training difficulties. To address these challenges a 3D dataset arranged in a 1D array format along with new architectures of 1D CapsNet is proposed. This effectively reduces the total number of parameters, resulting in faster and more accurate training of a neural network. Using the proposed architecture also appears to improve the training accuracy even when the datasets are relatively small. For this work, ModelNet-10 and the ModelNet-40 datasets are used. They have indoor home object images in 3D, which are few when compared to other datasets.

This paper outlines structures of different automatic speech recognition systems, hybrid and end-to-end, pros and cons for each of them, including the comparison of training data and computational resources requirements. Three main approaches to speech recognition are considered: hybrid Hidden Markov Model – Deep Neural Network, end-to-end Connectionist Temporal Classification and Sequence-to-Sequence. The Listen, Attend, and Spell approach is chosen as an example for the Sequence-to-Sequence model.

Sentiment analysis is the task of automatically identifying the sentiment expressed in text. It has become increasingly important in many applications such as social media monitoring, product reviews analysis, and customer feedback evaluation. With the advent of deep learning techniques, sentiment analysis has seen significant improvements in performance and accuracy. This paper presents a comprehensive survey of machine learning and deep learning methods for sentiment analysis at the document, sentence, and aspect levels. We first provide an overview of traditional machine learning approaches to sentiment analysis and their limitations. We then look into various machine learning and deep learning architectures that have been successfully applied to this task. Additionally, we discuss the challenges of dealing with different data modalities, such as visual and multimodal data, and how both techniques have been adapted to address these challenges. Furthermore, we explore the applications of sentiment analysis in diverse domains, including social media, product reviews, and healthcare. Finally, we highlight the current limitations of deep learning approaches for sentiment analysis and outline potential future research directions. This survey aims to provide researchers and practitioners with a comprehensive understanding of the state-of-the-art deep learning techniques for sentiment analysis and their practical applications.

A automacao agricola tornou-se significativa em nosso pais devido a necessidade das empresas nacionais de competirem de modo adequado com empresas estrangeiras e do aumento da produtividade e reducao de perdas. Com o uso de recursos de avancada tecnologia como dispositivos eletronicos embarcados e tecnicas de agricultura de precisao e controle de processos,ja e possivel o aumento da producao. As culturas horticolas sao de grande importância economica, social e alimentar para a populacao mundial. O grande desafio atual no setor de mecanizacao agricola de horticolas e a colheita.Para isso o desenvolvimento de um sistema de visao computacional com emprego de processamento de imagem e redes neurais artificiais que permita a identificacao de horticolas e forneca parâmetros de posicionamento a um conjunto atuador robotico, para operacoes de colheita, atenderia a este desafio. O objetivo do trabalho foi desenvolver um sistema de deteccao, que atraves destes recursos possibilite a deteccao ...

The rise of Automated Film Censorship and Rating (CBFC) in machine learning (ML) reflects the challenges posed by the proliferation of audiovisual content across various platforms. Manual censorship struggles to keep pace, prompting researchers to explore ML-based solutions. To develop effective CBFC systems, a robust conceptual model encompassing diverse content categories and criteria is essential. This paper aims to delineate both general and context-specific content categories while establishing CBFC criteria. It conducts a systematic review of contemporary CBFC systems, elucidating their ML models, strengths, and weaknesses. Comparative analysis underscores the efficacy of sequential and multimodal approaches in CBFC development. Amidst a constantly evolving film landscape, categorizing and classifying content accurately is increasingly complex. The integration of AI Masters, cutting-edge ML systems capable of analyzing extensive datasets and making informed decisions, emerges as a pivotal advancement in film censorship and ratings. This article delves into the transformative impact of AI Masters on the film industry, highlighting their remarkable capabilities in reshaping traditional practices.

A novel pre-training method is proposed to improve deep-neural-networks (DNN) and long-short-term-memory (LSTM) performance, and reduce the local minimum problem for speech enhancement. We propose initializing the last layer weights of DNN and LSTM by Non-Negative-Matrix-Factorization (NMF) basis transposed values instead of random weights. Due to its ability to extract speech features even in presence of non-stationary noises, NMF is faster and more successful than previous pre-training methods for network convergence. Using NMF basis matrix in the first layer along with another pre-training method is also proposed. To achieve better results, we further propose training individual models for each noise type based on a noise classification strategy. The evaluation of the proposed method on TIMIT data shows that it outperforms the baselines significantly in terms of perceptual-evaluation-of-speech-quality (PESQ) and other objective measures. Our method outperforms the baselines in terms of PESQ up to 0.17, with an improvement percentage of 3.4%.

A credit card is an important financial tool that has emerged in parallel with the developments in technology from the past to the present and has become an indispensable part of human life. The credit card has many advantages that can be listed as facilitating online shopping, providing installments in purchases, and preventing cash dependence. This is why the rate of use of credit cards worldwide is increasing day by day. On the other hand, there are some risks of the credit cards highlighted by security concerns. The fraudsters who access the identity and credit card information of the consumers through different means use it to shop online without the consumer’s knowledge and gain an unfair advantage. Therefore, it is crucial to eliminate this security vulnerability that the fraudsters exploit and to develop an effective solution to the customer victimization experienced by e-commerce companies due to the fraudulent credit card transactions. With this motivation, the performance...

Object perception is a fundamental sub-field of Computer Vision, covering a multitude of individual areas and having contributed high-impact results. While Machine Learning has been traditionally applied to address related problems, recent works also seek ways to integrate knowledge engineering in order to expand the level of intelligence of the visual interpretation of objects, their properties and their relations with their environment. In this paper, we attempt a systematic investigation of how knowledge-based methods contribute to diverse object perception tasks. We review the latest achievements and identify prominent research directions.

Deep learning is progressively gaining popularity as a viable alternative to i-vectors for speaker recognition. Promising results have been recently obtained with Convolutional Neural Networks (CNNs) when fed by raw speech samples directly. Rather than employing standard hand-crafted features, the latter CNNs learn low-level speech representations from waveforms, potentially allowing the network to better capture important narrow-band speaker characteristics such as pitch and formants. Proper design of the neural network is crucial to achieve this goal. This paper proposes a novel CNN architecture, called SincNet, that encourages the first convolutional layer to discover more meaningful filters. SincNet is based on parametrized sinc functions, which implement band-pass filters. In contrast to standard CNNs, that learn all elements of each filter, only low and high cutoff frequencies are directly learned from data with the proposed method. This offers a very compact and efficient way to derive a customized filter bank specifically tuned for the desired application. Our experiments, conducted on both speaker identification and speaker verification tasks, show that the proposed architecture converges faster and performs better than a standard CNN on raw waveforms.

This paper outlines structures of different automatic speech recognition systems, hybrid and end-to-end, pros and cons for each of them, including the comparison of training data and computational resources requirements. Three main approaches to speech recognition are considered: hybrid Hidden Markov Model – Deep Neural Network, end-to-end Connectionist Temporal Classification and Sequence-to-Sequence. The Listen, Attend, and Spell approach is chosen as an example for the Sequence-to-Sequence model.

The use of robotic systems in logistics has increased the importance of precise positioning, especially in warehouses. The paper presents a system that uses virtual fiducial markers to accurately predict the position of a drone in a warehouse and count items on the rack. A warehouse scenario is created in the simulation environment to determine the success rate of positioning. A total of 27 racks are lined up in the warehouse and in the center of the space, and a 6 × 6 ArUco type fiducial marker is used on each rack. The position of the vehicle is predicted by supervised learning. The inputs are the virtual fiducial marker features from the drone. The output data are the cartesian position and yaw angle. All input and output data required for supervised learning in the simulation environment were collected along different random routes. An image processing algorithm was prepared by making use of fiducial markers to perform rack counting after the positioning process. Among the regre...

Introduction: This paper presents Enactive Artificial Intelligence (eAI) as a genderinclusive approach to AI, emphasizing the need to address social marginalization resulting from unrepresentative AI design. Methods: The study employs a multidisciplinary framework to explore the intersectionality of gender and technoscience, focusing on the subversion of gender norms within Robot-Human Interaction in AI. Results: The results reveal the development of four ethical vectors, namely explainability, fairness, transparency, and auditability, as essential components for adopting an inclusive stance and promoting gender-inclusive AI. Discussion: By considering these vectors, we can ensure that AI aligns with societal values, promotes equity and justice, and facilitates the creation of a more just and equitable society. KEYWORDS artificial intelligence, human-robot interaction, feminist technoscience, enactivism, sociocultural practices. Introduction: male gaze in technoscience In recent years, there has been an unprecedented rise in automation, with Artificial Intelligence (AI) at the forefront of robotics trends. Collaborative robots, robot employees, and processes of automation in customer service, computer vision, and natural language processing are just some of the cutting-edge AI robotics trends

The proposed work presents a framework based on Graph Neural Networks (GNN) that abstracts the task to be executed and directly allows the robot to learn task-specific rules from synthetic demonstrations given through imitation learning. A graph representation of the state space is considered to encode the task-relevant entities as nodes for a Pick-and-Place task declined at different levels of difficulty. During training, the GNN-based policy learns the underlying rules of the manipulation task focusing on the structural relevance and the type of objects and goals, relying on an external primitive to move the robot to accomplish the task. The GNN-policy has been trained as a node-classification approach by looking at the different configurations of the objects and goals present in the scene, learning the association between them with respect to their type for the Pick-and-Place task. The experimental results show a high generalization capability of the proposed model in terms of the number, positions, height distributions, and even configurations of the objects/goals. Thanks to the generalization, only a single image of the desired goal configuration is required at inference time.

Offset quadrature amplitude modulation-based filter bank multicarrier (FBMC/OQAM) is among the promising waveforms for future wireless communication systems. This is due to its flexible spectrum usage and high spectral efficiency compared with the conventional multicarrier schemes. However, with OQAM modulation, the FBMC/OQAM signals are not orthogonal in the imaginary field. This causes a significant intrinsic interference, which is an obstacle to apply multiple input multiple output (MIMO) technology with FBMC/OQAM. In this paper, we propose a deep neural network (DNN)-based approach to deal with the imaginary interference, and enable the application of MIMO technique with FBMC/OQAM. We show, by simulations, that the proposed approach provides good performance in terms of bit error rate (BER).

Object perception is a fundamental sub-field of Computer Vision, covering a multitude of individual areas and having contributed high-impact results. While Machine Learning has been traditionally applied to address related problems, recent studies also seek ways to integrate knowledge engineering in order to expand the level of intelligence of the visual interpretation of objects, their properties and their relations with the environment. In this paper, we attempt a systematic investigation of how knowledge-based methods contribute to diverse object perception tasks. We review the latest achievements and identify prominent research directions.

In this paper, we introduce MFCCGAN as a novel speech synthesizer based on adversarial learning that adopts MFCCs as input and generates raw speech waveforms. Benefiting the GAN model capabilities, it produces speech with higher intelligibility than a rule-based MFCC-based speech synthesizer WORLD. We evaluated the model based on a popular intrusive objective speech intelligibility measure (STOI) and quality (NISQA score). Experimental results show that our proposed system outperforms Librosa MFCCinversion (by an increase of about 26% up to 53% in STOI and 16% up to 78% in NISQA score) and a rise of about 10% in intelligibility and about 4% in naturalness in comparison with conventional rule-based vocoder WORLD that used in the CycleGAN-VC family. However, WORLD needs additional data like F0. Finally, using perceptual loss in discriminators based on STOI could improve the quality more. WebMUSHRA-based subjective tests also show the quality of the proposed approach.

In the current industrial landscape, increasingly pervaded by technological innovations, the adoption of optimized strategies for asset management is becoming a critical key success factor. Among the various strategies available, the “Prognostics and Health Management” strategy is able to support maintenance management decisions more accurately, through continuous monitoring of equipment health and “Remaining Useful Life” forecasting. In the present study, convolutional neural network-based deep neural network techniques are investigated for the remaining useful life prediction of a punch tool, whose degradation is caused by working surface deformations during the machining process. Surface deformation is determined using a 3D scanning sensor capable of returning point clouds with micrometric accuracy during the operation of the punching machine, avoiding both downtime and human intervention. The 3D point clouds thus obtained are transformed into bidimensional image-type maps, i.e.,...

The automatic modulation classification (AMC) of a detected signal has gained considerable prominence in recent years owing to its numerous facilities. Numerous studies have focused on feature-based AMC. However, improving accuracy under low signal-to-noise ratio (SNR) rates is a serious issue in AMC. Moreover, research on the enhancement of AMC performance under low and high SNR rates is limited. Motivated by these issues, this study proposes AMC using a feature clustering-based two-lane capsule network (AMC2N). In the AMC2N, accuracy of the MC process is improved by designing a new two-layer capsule network (TL-CapsNet), and classification time is reduced by introducing a new feature clustering approach in the TL-CapsNet. Firstly, the AMC2N executes blind equalization, sampling, and quantization in trilevel preprocessing. Blind equalization is executed using a binary constant modulus algorithm to avoid intersymbol interference. To extract features from the preprocessed signal and ...

Trademarked names, logos, and images may appear in this book. Rather than use a trademark symbol with every occurrence of a trademarked name, logo, or image we use the names, logos, and images only in an editorial fashion and to the benefit of the trademark owner, with no intention of infringement of the trademark. The use in this publication of trade names, trademarks, service marks, and similar terms, even if they are not identified as such, is not to be taken as an expression of opinion as to whether or not they are subject to proprietary rights.

A credit card is an important financial tool that has emerged in parallel with the developments in technology from the past to the present and has become an indispensable part of human life. The credit card has many advantages that can be listed as facilitating online shopping, providing installments in purchases, and preventing cash dependence. This is why the rate of use of credit cards worldwide is increasing day by day. On the other hand, there are some risks of the credit cards highlighted by security concerns. The fraudsters who access the identity and credit card information of the consumers through different means use it to shop online without the consumer's knowledge and gain an unfair advantage. Therefore, it is crucial to eliminate this security vulnerability that the fraudsters exploit and to develop an effective solution to the customer victimization experienced by e-commerce companies due to the fraudulent credit card transactions. With this motivation, the performance of the methods from different research fields was examined to explore the solution space in detail in terms of the problem at hand within the scope of this study. For this purpose, three machine learning algorithms (K-Nearest Neighbor, Naive Bayes, Support Vector Machine), two artificial neural network algorithms (Binary Classifier, Autoencoder), and two deep learning algorithms (Deep Autoencoder and Deep Neural Network Classifier) were implemented. The effectiveness of the algorithms in question was tested with a famous dataset widely used in the literature. Experimental results showed that the Deep Neural Network Classifier outperformed the other algorithms used in this study and the best study ever reported in the literature in detecting fraudulent credit card transactions when accuracy and AUROC performance criteria were taken into account.

This paper introduces a novel method for determining the best room to place an object in, for embodied scene rearrangement. While state-of-the-art approaches rely on large language models (LLMs) or reinforcement learned (RL) policies for this task, our approach, CLIPGraphs, efficiently combines commonsense domain knowledge, data-driven methods, and recent advances in multimodal learning. Specifically, it (a) encodes a knowledge graph of prior human preferences about the room location of different objects in home environments, (b) incorporates vision-language features to support multimodal queries based on images or text, and (c) uses a graph network to learn object-room affinities based on embeddings of the prior knowledge and the vision-language features. We demonstrate that our approach provides better estimates of the most appropriate location of objects from a benchmark set of object categories in comparison with state-of-the-art baselines 1 .

State of the art algorithms for many pattern recognition problems rely on deep network models. Training these models requires a large labeled dataset and considerable computational resources. Also, it is difficult to understand the working of these learned models, limiting their use in some critical applications. Towards addressing these limitations, our architecture draws inspiration from research in cognitive systems, and integrates the principles of commonsense logical reasoning, inductive learning, and deep learning. In the context of answering explanatory questions about scenes and the underlying classification problems, the architecture uses deep networks for extracting features from images and for generating answers to queries. Between these deep networks, it embeds components for non-monotonic logical reasoning with incomplete commonsense domain knowledge, and for decision tree induction. It also incrementally learns and reasons with previously unknown constraints governing the domain's states. We evaluated the architecture in the context of datasets of simulated and real-world images, and a simulated robot computing, executing, and providing explanatory descriptions of plans. Experimental results indicate that in comparison with an "end to end" architecture of deep networks, our architecture provides better accuracy on classification problems when the training dataset is small, comparable accuracy with larger datasets, and more accurate answers to explanatory questions. Furthermore, incremental acquisition of previously unknown constraints improves the ability to answer explanatory questions, and extending non-monotonic logical reasoning to support planning and diagnostics improves the reliability and efficiency of computing and executing plans on a simulated robot.

Understanding how deep neural networks resemble or di er from human vision becomes increasingly important with their widespread use in Computer Vision and as models in Neuroscience. A key aspect of human vision is shape: we decompose the visual world into distinct objects, use cues to infer their D geometries, and can group several object parts into a coherent whole. Do deep networks use the shape of objects similarly when they classify images? Research on this question has yielded conflicting results, with some studies showing evidence for shape selectivity in deep networks, while others demonstrated clear deficiencies. We argue that these conflicts arise from di erences in experimental methods: whether studies use custom images in which only some features are available, images in which di erent features compete, image pairs that vary along di erent feature dimensions, or large sets of images to assess how representations vary overall. Each method o ers a di erent, partial view of shape processing. After comparing their advantages and pitfalls, we propose two hypotheses that can reconcile previous results. Firstly, deep networks are sensitive to local, but not global shape. Secondly, the higher layers of deep networks discard some of the shape information that the lower layers are sensitive to. We test these hypotheses by comparing network representations for natural images and silhouettes in which local or global shape is degraded. The results support both hypotheses, but for di erent networks. Purely feed-forward convolutional networks are unable to integrate shape globally. In contrast, networks with residual or recurrent connections show a weak selectivity for global shape. This motivates further research into recurrent architectures for perceptual integration.

This study developed a novel method for analyzing and decomposing a signal into its main dynamics for small and large timescales. Our proposal is based on a decoupled hybrid system of convolutional and recurrent neural networks that uses as inputs the power spectrum and spectrogram of a given signal, giving as output the dynamic behavior. We define the dynamic classification predicted of the signal using previously known dynamics characterized through training signals: periodic, quasi-periodic, aperiodic, chaotic, and randomness. We created a synthetic dataset comprising more than 50 training signals from different categories. For the real-world dataset, we used photoplethysmographic signals from 40 students obtained from a Spanish medical study. We tested the developed system's performance in real biological and synthetical signals, obtaining noteworthy results. All the results are evaluated qualitatively and quantitatively. Still, given the novelty and the lack of similar works, we cannot compare reliably and rigorously our results with other works, at least quantitatively. We can retrieve from the exposed results in this work three key ideas: the DNN-based solutions are capable of learning and generalizing the dynamics behavior of signals; the proposal learned correctly to distinguish between the reference dynamics provided and find some unidirectional similarities in the aperiodicity cases; and the results obtained using real-world PPG signals reveal that biological signals seem to exhibit a multi-dynamic behavior that changes depending on the used timescale, being quasi-periodically dominant in the short-term and aperiodically dominant in the long-term.

A credit card is an important financial tool that has emerged in parallel with the developments in technology from the past to the present and has become an indispensable part of human life. The credit card has many advantages that can be listed as facilitating online shopping, providing installments in purchases, and preventing cash dependence. This is why the rate of use of credit cards worldwide is increasing day by day. On the other hand, there are some risks of the credit cards highlighted by security concerns. The fraudsters who access the identity and credit card information of the consumers through different means use it to shop online without the consumer's knowledge and gain an unfair advantage. Therefore, it is crucial to eliminate this security vulnerability that the fraudsters exploit and to develop an effective solution to the customer victimization experienced by e-commerce companies due to the fraudulent credit card transactions. With this motivation, the performance of the methods from different research fields was examined to explore the solution space in detail in terms of the problem at hand within the scope of this study. For this purpose, three machine learning algorithms (K-Nearest Neighbor, Naive Bayes, Support Vector Machine), two artificial neural network algorithms (Binary Classifier, Autoencoder), and two deep learning algorithms (Deep Autoencoder and Deep Neural Network Classifier) were implemented. The effectiveness of the algorithms in question was tested with a famous dataset widely used in the literature. Experimental results showed that the Deep Neural Network Classifier outperformed the other algorithms used in this study and the best study ever reported in the literature in detecting fraudulent credit card transactions when accuracy and AUROC performance criteria were taken into account.

In this paper, we have created a system to detect whether a person is wearing a mask or not in public places like the mall, where we have CCTV cameras. The result can be shared with different screens so that others can prevent themself from that person or officials can charge the respective person. The system aims to prevent viral transmission in case of any pandemic.

The proposed work presents a framework based on Graph Neural Networks (GNN) that abstracts the task to be executed and directly allows the robot to learn task-specific rules from synthetic demonstrations given through imitation learning. A graph representation of the state space is considered to encode the task-relevant entities as nodes for a Pick-and-Place task declined at different levels of difficulty. During training, the GNN-based policy learns the underlying rules of the manipulation task focusing on the structural relevance and the type of objects and goals, relying on an external primitive to move the robot to accomplish the task. The GNN-policy has been trained as a node-classification approach by looking at the different configurations of the objects and goals present in the scene, learning the association between them with respect to their type for the Pick-and-Place task. The experimental results show a high generalization capability of the proposed model in terms of the number, positions, height distributions, and even configurations of the objects/goals. Thanks to the generalization, only a single image of the desired goal configuration is required at inference time.

During the image acquisition process, some level of noise is usually added to the real data mainly due to physical limitations of the acquisition sensor, and also regarding imprecisions during the data transmission and manipulation. Therefore, the resultant image needs to be processed in order to attenuate its noise without loosing details. Machine learning approaches have been successfully used for image denoising. Among such approaches, Restricted Boltzmann Machine (RBM) is one of the most used technique for this purpose. Here, we propose to enhance the RBM performance on image denoising by adding a posterior supervision before its final denoising step. To this purpose, we propose a simple but effective approach that performs a fine-tuning in the RBM model. Experiments on public datasets corrupted by different levels of Gaussian noise support the effectiveness of the proposed approach with respect to some state-of-the-art image denoising approaches.

The PSNR-oriented super-resolution (SR) methods pursue high reconstruction accuracy, but tend to produce over-smoothed results and lose plenty of high-frequency details. The GAN-based SR methods aim to generate more photo-realistic images, but the hallucinatory details are often accompanied with unsatisfying artifacts and noise. To address these problems, we propose a guided dual super-resolution network (GDSR), which exploits the advantages of both the PSNR-oriented and the GAN-based methods to achieve a good trade-off between reconstruction accuracy and perceptual quality. Specifically, our network contains two branches, where one is trained to extract global information and the other to focus on detail information. In this way, our network simultaneously generates SR images with high accuracy and satisfactory visual quality. To obtain more high-frequency features, we use the global features extracted from the low-frequency branch to guide the training of the high-frequency branch. Besides, our method utilizes a mask network to adaptively recover the final super-resolved image. Extensive experiments on several standard benchmarks show that our proposed method achieves better performance compared with state-of-the-art methods. The source code and the results of our GDSR are available at https://github.com/wenchen4321/GDSR.

Automatic Speaker Identification (ASI) is one of the active fields of research in signal processing. Various machine learning algorithms have been used for this purpose. With the recent developments in hardware technologies and data accumulation, Deep Learning (DL) methods have become the new state-of-the-art approach in several classification and identification tasks. In this paper, we evaluate the performance of traditional methods such as Gaussian Mixture Model-Universal Background Model (GMM-UBM) and DL-based techniques such as Factorized Time-Delay Neural Network (FTDNN) and Convolutional Neural Networks (CNN) for text-independent closed-set automatic speaker identification on two datasets with different conditions. LibriSpeech is one of the experimental datasets, which consists of clean audio signals from audiobooks, collected from a large number of speakers. The other dataset was collected and prepared by us, which has rather limited speech data with low signal-to-noise-ratio from real-life conversations of customers with the agents in a call center. The duration of the speech signals in the query phase is an important factor affecting the performances of ASI methods. In this work, a CNN architecture is proposed for automatic speaker identification from short speech segments. The architecture design aims at capturing the temporal nature of speech signal in an optimum convolutional neural network with low number of parameters compared to the well-known CNN architectures. We show that the proposed CNN-based algorithm performs better on the large and clean dataset, whereas on the other dataset with limited amount of data, traditional method outperforms all DL approaches. The achieved top-1 accuracy by the proposed model is 99.5% on 1-second voice instances from LibriSpeech dataset.