The growth in the volume of data generated, consumed, and stored, which is estimated to exceed 180 zettabytes in 2025, represents a major challenge both for organizations and for society in general. In addition to bei...
详细信息
The growth in the volume of data generated, consumed, and stored, which is estimated to exceed 180 zettabytes in 2025, represents a major challenge both for organizations and for society in general. In addition to being larger, datasets are increasingly complex, bringing new theoretical and computational challenges. Alongside this evolution, data science tools have exploded in popularity over the past two decades due to their myriad of applications when dealing with complex data, their high accuracy, flexible customization, and excellent adaptability. When it comes to images, data analysis presents additional challenges because as the quality of an image increases, which is desirable, so does the volume of data to be processed. Although classic machine learning (ML) techniques are still widely used in different research fields and industries, there has been great interest from the scientific community in the development of new artificial intelligence (AI) techniques. The resurgence of neuralnetworks has boosted remarkable advances in areas such as the understanding and processing of images. In this study, we conducted a comprehensive survey regarding advances in AI design and the optimization solutions proposed to deal with imageprocessing challenges. Despite the good results that have been achieved, there are still many challenges to face in this field of study. In this work, we discuss the main and more recent improvements, applications, and developments when targeting imageprocessingapplications, and we propose future research directions in this field of constant and fast evolution.
Graph neuralnetworks (GNNs) are neural models that use message transmission between graph nodes to represent the dependency of graphs. Variants of Graph neuralnetworks (GNNs), such as graph recurrent networks (GRN),...
详细信息
Graph neuralnetworks (GNNs) are neural models that use message transmission between graph nodes to represent the dependency of graphs. Variants of Graph neuralnetworks (GNNs), such as graph recurrent networks (GRN), graph attention networks (GAT), and graph convolutional networks (GCN), have shown remarkable results on a variety of deep learning tasks in recent years. In this study, we offer a generic design pipeline for GNN models, go over the variations of each part, classify the applications in an organized manner, and suggest four outstanding research issues. Dealing with graph data, which provides extensive connection information among pieces, is necessary for many learning tasks. A model that learns from graph inputs is required for modelling physics systems, learning molecular fingerprints, predicting protein interfaces, and identifying illnesses. Reasoning on extracted structures (such as the dependency trees of sentences and the scene graphs of photos) is an important research issue that also requires graph reasoning models in other domains, such as learning from non-structural data like texts and images. Graph neuralnetworks (GNNs) are primarily designed for dealing with graph-structured data, where relationships between entities are modeled as edges in a graph. While GNNs are not traditionally applied to image classification problems, researchers have explored ways to leverage graph-based structures to enhance the performance of Convolutional neuralnetworks (CNNs) in certain scenario. GNN have been increasingly applied to Natural Language processing (NLP) tasks, leveraging their ability to model structured data and capture relationships between elements in a graph. GNN are also applied for traffic related problems particularly in modeling and optimizing traffic flow, analyzing transportation networks, and addressing congestion issues. GNN can be used for traffic flow prediction, dynamic routing & navigation, Anomaly detection, public transport network
The Forward-Forward (FF) Algorithm has been recently proposed to alleviate the issues of backpropagation (BP) commonly used to train deep neuralnetworks. However, its current formulation exhibits limitations such as ...
详细信息
ISBN:
(纸本)1577358872
The Forward-Forward (FF) Algorithm has been recently proposed to alleviate the issues of backpropagation (BP) commonly used to train deep neuralnetworks. However, its current formulation exhibits limitations such as the generation of negative data, slower convergence, and inadequate performance on complex tasks. In this paper we take the main ideas of FF and improve them by leveraging channel-wise competitive learning in the context of convolutional neuralnetworks for image classification tasks. A layer-wise loss function is introduced that promotes competitive learning and eliminates the need for negative data construction. To enhance both the learning of compositional features and feature space partitioning, a channel-wise feature separator and extractor block is proposed that complements the competitive learning process. Our method outperforms recent FF-based models on image classification tasks, achieving testing errors of 0.58%, 7.69%, 21.89%, and 48.77% on MNIST, Fashion-MNIST, CIFAR-10 and CIFAR-100 respectively. Our approach bridges the performance gap between FF learning and BP methods, indicating the potential of our proposed approach to learn useful representations in a layer-wise modular fashion, enabling more efficient and flexible learning. Our source code and supplementary material are available at https://***/andreaspapac/CwComp.
This article explores the advancements and applications of Convolutional neuralnetworks (CNNs) in image classification, focusing on the CIFAR-10 dataset. Since their inception in 2006, CNNs have revolutionized comput...
详细信息
Recently, patch-wise contrastive learning is drawing attention for the image translation by exploring the semantic correspondence between the input and output images. To further explore the patch-wise topology for hig...
详细信息
ISBN:
(纸本)1577358872
Recently, patch-wise contrastive learning is drawing attention for the image translation by exploring the semantic correspondence between the input and output images. To further explore the patch-wise topology for high-level semantic understanding, here we exploit the graph neural network to capture the topology-aware features. Specifically, we construct the graph based on the patch-wise similarity from a pretrained encoder, whose adjacency matrix is shared to enhance the consistency of patch-wise relation between the input and the output. Then, we obtain the node feature from the graph neural network, and enhance the correspondence between the nodes by increasing mutual information using the contrastive loss. In order to capture the hierarchical semantic structure, we further propose the graph pooling. Experimental results demonstrate the state-of-art results for the image translation thanks to the semantic encoding by the constructed graphs.
With the continuous development of higher education in our country, the number of college students is increasing year by year. There are more and more campus accidents caused by college students' psychological pro...
详细信息
This article presents an artificial intelligence model capable of identifying actions strongly related to trichotillomania, a psychiatric disorder that causes people to have a desire to pull their hair. The model was ...
详细信息
Ensuring the secure and dependable deployment of deep neuralnetworks hinges on their ability to withstand distributional shifts and distortions. While data augmentation enhances robustness, its effectiveness varies a...
详细信息
ISBN:
(纸本)9798350349405;9798350349399
Ensuring the secure and dependable deployment of deep neuralnetworks hinges on their ability to withstand distributional shifts and distortions. While data augmentation enhances robustness, its effectiveness varies across different types of data corruption. It tends to excel in cases where corruptions share perceptually similar traits or have a high-frequency nature. In response, a strategy is to encompass a broad spectrum of distortions. Yet, it is often impractical to incorporate every conceivable modification that images may undergo within augmented data. Instead, we show that providing the model with a stronger inductive bias to learn the underlying concept of "change" would offer a more reliable approach. To this end, we develop Virtual Fusion (VF), a technique that treats corruptions as virtual labels. Diverging from conventional augmentation, when an image undergoes any form of transformation, its label becomes linked with the specific name attributed to the distortion. The finding indicates that VF effectively enhances both clean accuracy and robustness against common corruptions. On previously unseen corruptions, it shows an 11.90% performance improvement and a 12.78% increase in accuracy. In similar corruption scenarios, it achieves a 7.83% performance gain and a significant accuracy improvement of 22.04% on robustness benchmarks.
Structured pruning techniques have achieved great compression performance on convolutional neuralnetworks for image classification tasks. However, the majority of existing methods are sensitive with respect to the mo...
详细信息
ISBN:
(纸本)1577358872
Structured pruning techniques have achieved great compression performance on convolutional neuralnetworks for image classification tasks. However, the majority of existing methods are sensitive with respect to the model parameters, and their pruning results may be unsatisfactory when the original model is trained poorly. That is, they need the original model to be fully trained, to obtain useful weight information. This is time-consuming, and makes the effectiveness of the pruning results dependent on the degree of model optimization. To address the above issue, we propose a novel metric named Average Filter Information Entropy (AFIE). It decomposes the weight matrix of each layer into a low-rank space, and quantifies the filter importance based on the distribution of the normalized eigenvalues. Intuitively, the eigenvalues capture the covariance among filters, and therefore could be a good guide for pruning. Since the distribution of eigenvalues is robust to the updating of parameters, AFIE can yield a stable evaluation for the importance of each filter no matter whether the original model is trained fully. We implement our AFIE-based pruning method for three popular CNN models of AlexNet, VGG-16, and ResNet-50, and test them on three widely-used image datasets MNIST, CIFAR-10, and imageNet, respectively. The experimental results are encouraging. We surprisingly observe that for our methods, even when the original model is trained with only one epoch, the AFIE score of each filter keeps identical to the results when the model is fully-trained. This fully indicates the effectiveness of the proposed pruning method.
Digital security in modern systems very often uses biometric, and increasingly, new implementations appear. Such applications can be found everywhere, even when picking up the package from courier, we certify its rece...
详细信息
Digital security in modern systems very often uses biometric, and increasingly, new implementations appear. Such applications can be found everywhere, even when picking up the package from courier, we certify its receipt through our signature on the tablet. However, verification of this form is not one of the simplest elements in information processing systems. Given the different sizes, angles, or writing conditions that may affect its stability, new methods to evaluate signatures are constantly needed. In this article, we propose the use of spline interpolation and two types of artificialneuralnetworks to verify the identity of a person based on selected local and global features extracted from the image of a signature. Global features are extracted concerning interpolation and graphic processing methods, while local features are verified using convolutional neuralnetworks. Both sets of features are used in the identity verification process. The article presents the model of the operation together with experiments, taking into account various parameters of the proposed extraction. We have reached an accuracy of 87.7% on the SVC2004 database.
暂无评论