Image Classification Problems Research Articles

Implementation of satelite image classification using PyTorch

This article presents a practical implementation of the satellite image classification task using the PyTorch deep learning library. The aim of the work is to develop a neural network model capable of classifying satellite images into four classes: water, sands, clouds, and greenery. The model is trained using the pre-trained ResNet architecture. The proposed approach allows for automatic classification of satellite images, fa-cilitating the analysis of large amounts of data, and has been tested on Sentinel-2 satellite images acquired from March to August 2023 for parts of Kherson and Mykolaiv regions, Ukraine, and achieved an average accuracy of 95%. The presented research demonstrates the practical use of modern deep learning methods, in particular the PyTorch library, to solve actual problems of satellite image analysis and classification.

Two novel two-step inertial algorithms for a class of bilevel variational inequalities

This paper introduces two novel two-step inertial and self-adaptive step sizes algorithms for solving a strongly monotone variational inequality over the solution set of a split convex feasibility problem with multiple output sets in real Hilbert spaces. We establish strong convergence properties of the proposed method under mild conditions and employ it to solve monotone variational inequalities over the solution set of the split feasibility problem. The method is further applied to the image classification problem via the support vector machine learning. Numerical results are given to demonstrate the accelerating behaviors of our method over other related methods in the literature.

A Copula Based Unsupervised Domain Adaptation for Image Classification

In this paper, we present an unsupervised domain adaptation algorithm for image classification using principal component analysis (PCA) and Gaussian copula function alignment. The motivation of the proposed algorithm stems from the idea of CORAL algorithm which extracts domain invariant features by aligning the correlation structure between a source and a target domain. However, it suffers from the fact that highly skewed marginal distributions happen to distort the correlation structure so that it may cause a negative transfer. Therefore we utilize a copula function that enables us to analyze separately the dependency structure and the marginals by Sklar’s theorem. In particular, we propose to align the Gaussian copula correlations in the copula feature spaces instead of aligning the correlation matrices in the original space. Considering the extremely skewed distribution of SURF image features in Office-Caltech10 data set we apply PCA first in order to extract some skewness-mitigated principal features and then derive copula features to align for domain adaptation by using CORAL idea with Gaussian copula correlation matrices. The proposed method showed a good classification accuracy when applied to image classification problem in an unsupervised domain adaptation setting.

Intelligent agent for hyperspectral image classification with noisy labels: a deep reinforcement learning framework

ABSTRACT Label accuracy plays an important role in supervised hyperspectral image classification problems. However, labelling can be prone to various types of noise, mainly including human errors and incorrect pseudo-label predictions. Learning from datasets with noisy labels is a critical challenge in practice. To address this issue, we first propose a deep reinforcement learning method named HyperRefine to tackle the label noise problem for hyperspectral image classification. The method utilizes an agent to learn the accuracy of each sample label and determine its suitability to be used in training the classifier model. The agent is trained using a reinforcement signal based on the reward obtained from a small validation set that reflects performance on the target task, where the validation set is generated from the training set by a momentum-based recruiter. Compared to complex hand-crafted methods of label noise cleaning, HyperRefine is easy to implement, yet highly effective. The robustness of the model to different types of label noises is significantly improved by further combining a novel label refinement approach. Experiments on artificially added label noise and pseudo-label noise hyperspectral datasets show that the proposed method can achieve promising performance.

PlasmoData.jl — A Julia framework for modeling and analyzing complex data as graphs

Datasets encountered in scientific and engineering applications appear in complex formats (e.g., images, multivariate time series, molecules, video, text strings, networks). Graph theory provides a unifying framework to model such datasets and enables the use of powerful tools that can help analyze, visualize, and extract value from data. In this work, we present PlasmoData.jl, an open-source, Julia framework that uses concepts of graph theory to facilitate the modeling and analysis of complex datasets. The core of our framework is a general data modeling abstraction, which we call a DataGraph. We show how the abstraction and software implementation can be used to represent diverse data objects as graphs and to enable the use of tools from topology, graph theory, and machine learning (e.g., graph neural networks) to conduct a variety of tasks. We illustrate the versatility of the framework by using real datasets: (i) an image classification problem using topological data analysis to extract features from the graph model to train machine learning models; (ii) a disease outbreak problem where we model multivariate time series as graphs to detect abnormal events; and (iii) a technology pathway analysis problem where we highlight how we can use graphs to navigate connectivity. Our discussion also highlights how PlasmoData.jl leverages native Julia capabilities to enable compact syntax, scalable computations, and interfaces with diverse packages. Overall, we show that the DataGraph abstraction and PlasmoData.jl Julia package are able to model data within graphs and enable useful analysis.

Self-reconstruction network for fine-grained few-shot classification

Metric-based methods are one of the most common methods to solve the problem of few-shot image classification. However, traditional metric-based few-shot methods suffer from overfitting and local feature misalignment. The recently proposed feature reconstruction-based approach, which reconstructs query image features from the support set features of a given class and compares the distance between the original query features and the reconstructed query features as the classification criterion, effectively solves the feature misalignment problem. However, the issue of overfitting still has not been considered. To this end, we propose a self-reconstruction metric module for diversifying query features and a restrained cross-entropy loss for avoiding over-confident predictions. By introducing them, the proposed self-reconstruction network can effectively alleviate overfitting. Extensive experiments on five benchmark fine-grained datasets demonstrate that our proposed method achieves state-of-the-art performance on both 5-way 1-shot and 5-way 5-shot classification tasks. Code is available at https://github.com/liz-lut/SRM-main.

A lightweight convolutional neural network (CNN) model for diatom classification: DiatomNet.

Diatoms are a type of algae with many species. Accurate and quick classification of diatom species is important in many fields, such as water quality analysis and weather change forecasting. Traditional methods for diatom classification, specifically morphological taxonomy and molecular detection, are time-consuming and may not provide satisfactory performance. However, in recent years, deep learning has demonstrated impressive performance in this task, just like other image classification problems. On the other hand, networks with more layers do not guarantee increased accuracy. While increasing depth can be useful in capturing complex features and patterns, it also introduces challenges such as vanishing gradients, overfitting, and optimization challenges. Therefore, in our work, we propose DiatomNet, a lightweight convolutional neural network (CNN) model that can classify diatom species accurately while requiring low computing resources. A recently introduced dataset consisting of 3,027 diatom images and 68 diatom species is used to train and evaluate the model. The model is compared with well-known and successful CNN models (i.e., AlexNet, GoogleNet, Inceptionv3, ResNet18, VGG16, and Xception) and their customized versions obtained with transfer learning. The comparison is based on several success metrics: accuracy, precision, recall, F-measure, number of learnable parameters, training, and prediction time. Eventually, the experimental results reveal that DiatomNet outperforms the other models regarding all metrics with just a few exceptions. Therefore, it is a lightweight but strong candidate for diatom classification tasks.

A novel breast cancer image classification model based on multiscale texture feature analysis and dynamic learning

Assistive medical image classifiers can greatly reduce the workload of medical personnel. However, traditional machine learning methods require large amounts of well-labeled data and long learning times to solve medical image classification problems, which can lead to high training costs and poor applicability. To address this problem, a novel unsupervised breast cancer image classification model based on multiscale texture analysis and a dynamic learning strategy for mammograms is proposed in this paper. First, a gray-level cooccurrence matrix and Tamura coarseness are used to transfer images to multiscale texture feature vectors. Then, an unsupervised dynamic learning mechanism is used to classify these vectors. In the simulation experiments with a resolution of 40 pixels, the accuracy, precision, F1-score and AUC of the proposed method reach 91.500%, 92.780%, 91.370%, and 91.500%, respectively. The experimental results show that the proposed method can provide an effective reference for breast cancer diagnosis.

Adaptive habitat biogeography-based optimizer for optimizing deep CNN hyperparameters in image classification

Deep Convolutional Neural Networks (DCNNs) have shown remarkable success in image classification tasks, but optimizing their hyperparameters can be challenging due to their complex structure. This paper develops the Adaptive Habitat Biogeography-Based Optimizer (AHBBO) for tuning the hyperparameters of DCNNs in image classification tasks. In complicated optimization problems, the BBO suffers from premature convergence and insufficient exploration. In this regard, an adaptable habitat is presented as a solution to these problems; it would permit variable habitat sizes and regulated mutation. Better optimization performance and a greater chance of finding high-quality solutions across a wide range of problem domains are the results of this modification's increased exploration and population diversity. AHBBO is tested on 53 benchmark optimization functions and demonstrates its effectiveness in improving initial stochastic solutions and converging faster to the optimum. Furthermore, DCNN-AHBBO is compared to 23 well-known image classifiers on nine challenging image classification problems and shows superior performance in reducing the error rate by up to 5.14%. Our proposed algorithm outperforms 13 benchmark classifiers in 87 out of 95 evaluations, providing a high-performance and reliable solution for optimizing DNNs in image classification tasks. This research contributes to the field of deep learning by proposing a new optimization algorithm that can improve the efficiency of deep neural networks in image classification.

Construction of fuzzy classification systems by fitness sharing based genetic search and boosting based ensemble

This paper concentrates on the development of precise fuzzy rule-based classification systems for high-dimensional and multi-class problems. The approach begins with the extraction on potential fuzzy if-then rules using fitness sharing based genetic algorithms, this ensures effective searching for productive niches, thereby evolving and maintaining a diverse, cooperative population. Subsequently, for the purpose of combining the obtained fuzzy rules and eliminating their conflicts, an adaboost ensemble method is utilized, enhancing the accuracy of the fuzzy classification systems.Experiments have been conducted on 10 UCI datasets and 3 well-known image classification problems. The features for these image tasks were derived from the activation values of the final convolutional layer in pre-trained convolutional neural networks. These datasets, which were chosen to evaluate the effectiveness of the proposed approach, exhibit significant variation in terms of dimensionality and the number of class labels. Comparative analyses are carried out with conventional fuzzy rule-based classification methods, and the results demonstrate that the classification systems can be developed for complex problems, while maintaining a high-level of prediction accuracy.

A Memory-Efficient Neural Ordinary Differential Equation Framework Based on High-Level Adjoint Differentiation

Neural ordinary differential equations (neural ODEs) have emerged as a novel network architecture that bridges dynamical systems and deep learning. However, the gradient obtained with the continuous adjoint method in the vanilla neural ODE is not reverse-accurate. Other approaches suffer either from an excessive memory requirement due to deep computational graphs or from limited choices for the time integration scheme, hampering their application to large-scale complex dynamical systems. To achieve accurate gradients without compromising memory efficiency and flexibility, we present a new neural ODE framework, PNODE, based on high-level discrete adjoint algorithmic differentiation. By leveraging discrete adjoint time integrators and advanced checkpointing strategies tailored for these integrators, PNODE can provide a balance between memory and computational costs, while computing the gradients consistently and accurately. We provide an open-source implementation based on PyTorch and PETSc, one of the most commonly used portable, scalable scientific computing libraries. We demonstrate the performance through extensive numerical experiments on image classification and continuous normalizing flow problems. We show that PNODE achieves the highest memory efficiency when compared with other reverse-accurate methods. On the image classification problems, PNODE is up to two times faster than the vanilla neural ODE and up to 2.3 times faster than the best existing reverse-accurate method. We also show that PNODE enables the use of the implicit time integration methods that are needed for stiff dynamical systems.

Self-Localization of Guide Robots Through Image Classification

The field of autonomous robotic systems has advanced tremendously in the last few years, allowing them to perform complicated tasks in various contexts. One of the most important and useful applications of guide robots is the support of the blind. The successful implementation of this study requires a more accurate and powerful self-localization system for guide robots in indoor environments. This paper proposes a self-localization system for guide robots. To successfully implement this study, images were collected from the perspective of a robot inside a room, and a deep learning system such as a convolutional neural network (CNN) was used. An image-based self-localization guide robot image-classification system delivers a more accurate solution for indoor robot navigation. The more accurate solution of the guide robotic system opens a new window of the self-localization system and solves the more complex problem of indoor robot navigation. It makes a reliable interface between humans and robots. This study successfully demonstrated how a robot finds its initial position inside a room. A deep learning system, such as a convolutional neural network, trains the self-localization system as an image classification problem. The robot was placed inside the room to collect images using a panoramic camera. Two datasets were created from the room images based on the height above and below the chest. The above-mentioned method achieved a localization accuracy of 98.98%.

IntroUNET: Identifying introgressed alleles via semantic segmentation.

A growing body of evidence suggests that gene flow between closely related species is a widespread phenomenon. Alleles that introgress from one species into a close relative are typically neutral or deleterious, but sometimes confer a significant fitness advantage. Given the potential relevance to speciation and adaptation, numerous methods have therefore been devised to identify regions of the genome that have experienced introgression. Recently, supervised machine learning approaches have been shown to be highly effective for detecting introgression. One especially promising approach is to treat population genetic inference as an image classification problem, and feed an image representation of a population genetic alignment as input to a deep neural network that distinguishes among evolutionary models (i.e. introgression or no introgression). However, if we wish to investigate the full extent and fitness effects of introgression, merely identifying genomic regions in a population genetic alignment that harbor introgressed loci is insufficient-ideally we would be able to infer precisely which individuals have introgressed material and at which positions in the genome. Here we adapt a deep learning algorithm for semantic segmentation, the task of correctly identifying the type of object to which each individual pixel in an image belongs, to the task of identifying introgressed alleles. Our trained neural network is thus able to infer, for each individual in a two-population alignment, which of those individual's alleles were introgressed from the other population. We use simulated data to show that this approach is highly accurate, and that it can be readily extended to identify alleles that are introgressed from an unsampled "ghost" population, performing comparably to a supervised learning method tailored specifically to that task. Finally, we apply this method to data from Drosophila, showing that it is able to accurately recover introgressed haplotypes from real data. This analysis reveals that introgressed alleles are typically confined to lower frequencies within genic regions, suggestive of purifying selection, but are found at much higher frequencies in a region previously shown to be affected by adaptive introgression. Our method's success in recovering introgressed haplotypes in challenging real-world scenarios underscores the utility of deep learning approaches for making richer evolutionary inferences from genomic data.

Deep Neural Network for the Detection and Classification of Spontaneous Abortion Associated with Cervical Cancer

Cervical cancer is the most prevalent cancer and serious condition that currently poses a threat to the health of women. Automated cervical cancer diagnosis and classification from pap smear images has become essential due to the precision, dependency, and timely analysis of the cancer progression. The chance of occurrence of Spontaneous Abortion (SA) associated with cervical carcinoma can be identified from the pap smear image analysis. Advancements in Deep Neural Network (DNN) paved path to the efficient analysis of pap smear images for the detection of SA. The DNN must be optimized to obtain better accuracy in multiclass Herlev dataset pap smear image classification problem. The class carcinoma-in-situ is associated with SA and this class need to be identified without any error. For quantitative analysis of cell images, accurate localization of cell nuclei as well as the extraction of textural features are accomplished. Classification and decision-making are carried out as part of high-level processing to categorise the cancer phases into six classes. The domain aspects of cervical cancer are used to explore and present appropriate methods and techniques. This work focuses on the development of optimized DNN that can differentiate cancer classes in an efficient way. For Herlev images the proposed system provided multiclass average accuracy of 96.49%. Proposed EL-CNN architecture provided better outcomes than state-of-the-art systems.

ADEMNET architecture: An innovative solution for adaptive multi-class balancing problem in image classification

<span>In the field of medical image processing, achieving high performance in the classification of four types of dementia poses a significant challenge. This research presents a novel approach that outperforms existing methodologies, bringing about a transformative impact in this specialized domain. The method integrates the adaptive synthetic–nominal (ADASYN) technique with a DEMNET framework, resulting in a substantial performance improvement of 95.45% compared to current benchmarks. Through meticulous experimentation on a dementia dataset encompassing four distinct types, we consistently demonstrate significant enhancements achieved by the refined strategy. This innovation not only raises the performance standard but also provides a robust and adaptable solution that can be easily integrated into existing systems. The implications of this advancement open up new avenues for both research and practical applications. This work exemplifies the power of innovative approaches to push the limits of performance and establishes a new benchmark for excellence within this specific domain.</span>

ACANet: A Fine-grained Image Classification Optimization Method Based on Convolution and Attention Fusion

<p>The key to solve the problem of fine-grained image classification is to find the differentiation regions related to fine-grained features. In this paper, we try to add new network components to the original network and adjust various parameters to try to propose a new fine-grained image classification network. We propose a fine-grained image classification network based on the fusion of asymmetric convolution, convolution and self-attention mechanisms. Firstly, an enhanced module using asymmetric convolution to assist classical convolution proposed to help convolution learn deep features. Secondly, according to the common points of convolution and self-attention mechanism, we invented a fusion module of convolution and self-attention mechanism to improve the learning ability of the network.We integrate these two modules into the residual network and invent a new residual network .Finally, according to the experience, we design a new downsampling layer to adapt to the new component of the attention mechanism and improve the performance of the model. The experiment test on three publicly available datasets, and three methods for comparison. The results show that the new structure can effectively complete the task of fine-grained image classification, and the classification accuracy of different methods and different datasets are significantly improved.</p> <p> </p>

Fault diagnosis using signal processing and deep learning-based image pattern recognition

Abstract The vibration signal is a typical non-stationary signal, making it challenging to use traditional time-frequency analysis techniques for fault diagnosis. Therefore, this work investigates the processing of vibration signals and proposes a deep learning method based on processed signals for the fault diagnosis of ball bearings. In this work, the fault diagnosis is formulated as an image classification problem and solved with deep learning networks. The intrinsic mode functions (IMFs), converted from the vibration signals in the time domain, are then transformed into symmetrized dot pattern (SDP) images. In order to increase classification accuracy, the SDP parameters in this study are chosen by optimizing image similarity. The feasibility and accuracy of the proposed approach are examined experimentally.

Post-disaster building damage detection using multi-source variational domain adaptation

Timely detection of building damage after a disaster can provide support and help in saving lives and reducing losses. The emergence of transfer learning can solve the problem of difficulty in obtaining several labeled samples to train deep models. However, some degree of differences exists among different scenarios, which may affect the transfer performance. Furthermore, in reality, data can be collected from multiple historical scenarios but cannot be directly combined using single-source domain adaptation methods. Therefore, this study proposes a multi-source variational domain adaptation (MVDA) method to complete the task of post-disaster building assessment. The MVDA method consists of two stages: first, the distributions of each pair of source and target domains in specific feature spaces are aligned separately; second, the outputs of the pre-trained classifiers are aligned using domain-specific decision boundaries. This method maximizes the relevant information in the historical scene, solves the problem of inconsistent image classification in the current scene, and improves the migration efficiency from the history to the current disaster scene. The proposed approach is validated by two challenging multi-source transfer tasks using the post-disaster hurricane datasets. The average accuracy rate of 83.3% for the two tasks is achieved, obtaining an improvement of 0.9% compared with the state-of-the-art methods.

Learning Location From Shared Elevation Profiles in Fitness Apps: A Privacy Perspective

The extensive use of smartphones and wearable devices has facilitated many useful applications. For example, with Global Positioning System (GPS)-equipped smart and wearable devices, many applications can gather, process, and share rich metadata, such as geolocation, trajectories, elevation, and time. For example, fitness applications, such as Runkeeper and Strava, utilize the information for activity tracking and have recently witnessed a boom in popularity. Those fitness tracker applications have their own web platforms and allow users to share activities on such platforms or even with other social network platforms. To preserve the privacy of users while allowing sharing, several of those platforms may allow users to disclose partial information, such as the elevation profile for an activity, which supposedly would not leak the location of the users. In this work, and as a cautionary tale, we create a proof of concept where we examine the extent to which elevation profiles can be used to predict the location of users. To tackle this problem, we devise three plausible threat settings under which the city or borough of the targets can be predicted. Those threat settings define the amount of information available to the adversary to launch the prediction attacks. Establishing that simple features of elevation profiles, e.g., spectral features, are insufficient, we devise both natural language processing (NLP)-inspired text-like representation and computer vision-inspired image-like representation of elevation profiles, and we convert the problem at hand into text and image classification problem. We use both traditional machine learning- and deep learning-based techniques and achieve a prediction success rate ranging from 59.59% to 99.80%. The findings are alarming, highlighting that sharing elevation information may have significant location privacy risks.

Detection and Classification of MRI Images using Multistage Classifier for Early Prediction of Alzheimer’s Disease

The term “neuro imaging” describes a set of procedures used for studying, diagnosing, and treating neurological disorders as well as for basic scientific research into the human brain. In order for medical professionals to make early diagnoses for their patients, they must investigate the signs of various neuro imaging kinds; among the most important elements in this process is the problem of neuro image classification. No effective treatment has been found for Alzheimer's disease (AD), a degenerative brain ailment that has no known cure. It is not possible to halt the progression of the disease once it has begun. But there are medications that can slow it down. Nevertheless, with the use of sophisticated prediction, the disease-influencing protein functions can be reduced. Because of shared neural networks and pixel strength, Alzheimer's disease diagnosis in the elderly is discretely challenging and calls for the portrayal of a discriminating element separately. Several writers attempted to apply machine learning to make early disease diagnoses, but they were unable to get reliable classification accuracy. To that end, we set out to develop a model that would combine deep neural networks with multistage classifiers in the hopes that it would be able to efficiently and effectively extract characteristics from input data. Better and more effective detection and classification of Alzheimer's disease was achieved in this study by employing a multistage classifier that made use of deep learning techniques. Using a benchmark database supplied by the Alzheimer's Neuro Imaging Institute, the suggested strategy achieves better outcomes than individual techniques.