Accuracy Of Image Classification Research Articles

Novel Class Discovery in Chest X-rays via Paired Images and Text

Novel class discover(NCD) aims to identify new classes undefined during model training phase with the help of knowledge of known classes. Many methods have been proposed and notably boosted performance of NCD in natural images. However, there has been no work done in discovering new classes based on medical images and disease categories, which is crucial for understanding and diagnosing specific diseases. Moreover, most of the existing methods only utilize information from image modality and use labels as the only supervisory information. In this paper, we propose a multi-modal novel class discovery method based on paired images and text, inspired by the low classification accuracy of chest X-ray images and the relatively higher accuracy of the paired text. Specifically, we first pretrain the image encoder and text encoder with multi-modal contrastive learning on the entire dataset and then we generate pseudo-labels separately on the image branch and text branch. We utilize intra-modal consistency to assess the quality of pseudo-labels and adjust the weights of the pseudo-labels from both branches to generate the ultimate pseudo-labels for training. Experiments on eight subset splits of MIMIC-CXR-JPG dataset show that our method improves the clustering performance of unlabeled classes by about 10% on average compared to state-of-the-art methods. Code is available at: https://github.com/zzzzzzzzjy/MMNCD-main.

Concatenated Modified LeNet Approach for Classifying Pneumonia Images.

Pneumonia remains a critical health concern worldwide, necessitating efficient diagnostic tools to enhance patient care. This research proposes a concatenated modified LeNet classifier to classify pneumonia images accurately. The model leverages deep learning techniques to improve the diagnosis of Pneumonia, leading to more effective and timely treatment. Our modified LeNet architecture incorporates a revised Rectified Linear Unit (ReLU) activation function. This enhancement aims to boost the discriminative capacity of the features learned by the model. Furthermore, we integrate batch normalization to stabilize the training process and enhance performance within smaller, less complex, CNN architectures like LeNet. Batch normalization addresses internal covariate shift, a phenomenon where the distribution of activations within a network alter during training. These modifications help to prevent overfitting and decrease computational time. A comprehensive dataset is used to evaluate the model's performance, and the model is benchmarked against relevant deep-learning models. The results demonstrate a high recognition rate, with an accuracy of 96% in pneumonia image recognition. This research suggests that the Concatenated Modified LeNet classifier has the potential to be a highly useful tool for medical professionals in the diagnosis of pneumonia. By offering accurate and efficient image classification, our model could contribute to improved treatment decisions and patient outcomes.

Multiscale attention for few‐shot image classification

AbstractIn recent years, the application of traditional deep learning methods in the agricultural field using remote sensing techniques, such as crop area and growth monitoring, crop classification, and agricultural disaster monitoring, has been greatly facilitated by advancements in deep learning. The accuracy of image classification plays a crucial role in these applications. Although traditional deep learning methods have achieved significant success in remote sensing image classification, they often involve convolutional neural networks with a large number of parameters that require extensive optimization using numerous remote sensing images for training purposes. To address these challenges, we propose a novel approach called multiscale attention network (MAN) for sample‐based remote sensing image classification. This method consists primarily of feature extractors and attention modules to effectively utilize different scale features through multiscale feature training during the training phase. We evaluate our proposed method on three datasets comprising agricultural remote sensing images and observe superior performance compared to existing approaches. Furthermore, we validate its generalizability by testing it on an oil well indicator diagram specifically designed for classification tasks.

Combining audio and non-audio inputs in evolved neural networks for Ovenbird classification

ABSTRACT In the last several years, the use of neural networks as tools to automate species classification from digital data has increased. This has been due in part to the high classification accuracy of image classification through Convolutional Neural Networks (CNN). In the case of audio data, CNN-based recognisers are used to automate the classification of species in audio recordings by using information from sound visualisation (i.e. spectrograms). It is common for these recognisers to use the spectrogram as their sole input. However, researchers have other, non-audio data, such as habitat preferences of a species, phenology, and range information, which could improve species classification. We present how a single-species recogniser neural network’s accuracy can be improved by using non-audio data as inputs in addition to spectrogram information. We analyse the cause of the improvements: are they a result of having a neural network with a higher number of parameters or is it due to the use of the two inputs? We find that networks that use the two different inputs have a higher classification accuracy. This suggests that the accuracy of classifiers can be improved by giving them non-audio information about the location and conditions where the recordings were obtained.

Enhancing Image Classification Accuracy Based on AlexNet

This study explores the potential of deep learning for small-scale image classification tasks through the utilization of the classical AlexNet model on the CIFAR-10 dataset. The methodology involves meticulous architectural adjustments, comprehensive data preprocessing, and strategic optimization of the learning rate decay, resulting in substantial improvements in model performance. In addition, innovative data augmentation techniques are introduced to enhance the model's robustness and generalization capabilities. The experimental outcomes unequivocally underscore the efficacy of deep learning in addressing small-scale image classification challenges, offering versatile applications across diverse domains such as image recognition, autonomous driving, and medical diagnostics. Acknowledging the dynamic nature of deep learning, ongoing research endeavors are warranted. Future directions may encompass the exploration of more intricate neural network architectures, advanced data augmentation methodologies, and the implementation of interpretability tools to augment both performance and comprehensibility. This study provides compelling evidence of deep learning's aptitude for small-scale image classification tasks, offering valuable insights and guidance for future research and practical implementations. As the field of deep learning continues to evolve, this work aims to serve as a valuable reference and catalyst for further advancements in the domain.

Classification of optical coherence tomography images using deep machine learning methods

Backgraund. Optical coherence tomography (OCT) is a modern high-tech and informative method for detecting pathology of the retina and preretinal layers of the vitreous body. However, the description and interpretation of the research results require high qualifications and special training of an ophthalmologist, and significant time expenditure for the doctor and the patient. At the same time, the use of mathematical models based on artificial neural networks (ANN- models) currently makes it possible to automate many processes associated with image processing. Therefore, solving problems associated with automating the process of classifying OCT images based on ANN models is actual.
 Aims. To develop architectures of mathematical (computer) models based on deep learning of convolutional neural networks (CNN) for classification of OCT images of the retina. To compare the results of computational experiments conducted using Python tools in the Google Colaboratory with single-model and multi-model approaches and evaluate classification accuracy. To make conclusions about the optimal architecture of ANN models and the values of the hyperparameters used.
 Materials and methods. The original dataset, which was anonymized OCT images of real patients, included more than 2000 images obtained directly from the device in a resolution of 1920 × 969 × 24 BPP. The number of image classes is 12. To create the training and validation data sets, a subject area of 1100 × 550 × 24 BPP was “cut out.” Various approaches were studied: the possibility of using pretrained CNNs with transfer learning, techniques for resizing and augmenting images, as well as various combinations of hyperparameters of ANN-models. When compiling the model, the following parameters were used: Adam optimizer, categorical_crossentropy loss function, accuracy metric. All technological processes with images and ANN-models were carried out using Python language tools in Google Colaboratory.
 Results. Single-model and multi-model principles for classifying OCT images of the retina are proposed. Computational experiments on automated classification of such images obtained from a DRI OCT Triton 3D tomograph using various ANN model architectures showed an accuracy of 98-100% during training and validation and 85% during an additional test, which is a satisfactory result. The optimal architecture of the ANN model - a six-layer convolutional network - was selected and the values of its hyperparameters were determined.
 Conclusions. The results of deep training of convolutional neural network models with various architectures, their validation and testing showed satisfactory classification accuracy of retinal OCT images. These developments can be used in decision support systems in the field of ophthalmology.

Hybrid Quantum Image Classification and Federated Learning for Hepatic Steatosis Diagnosis.

In the realm of liver transplantation, accurately determining hepatic steatosis levels is crucial. Recognizing the essential need for improved diagnostic precision, particularly for optimizing diagnosis time by swiftly handling easy-to-solve cases and allowing the expert time to focus on more complex cases, this study aims to develop cutting-edge algorithms that enhance the classification of liver biopsy images. Additionally, the challenge of maintaining data privacy arises when creating automated algorithmic solutions, as sharing patient data between hospitals is restricted, further complicating the development and validation process. This research tackles diagnostic accuracy by leveraging novel techniques from the rapidly evolving field of quantum machine learning, known for their superior generalization abilities. Concurrently, it addresses privacy concerns through the implementation of privacy-conscious collaborative machine learning with federated learning. We introduce a hybrid quantum neural network model that leverages real-world clinical data to assess non-alcoholic liver steatosis accurately. This model achieves an image classification accuracy of 97%, surpassing traditional methods by 1.8%. Moreover, by employing a federated learning approach that allows data from different clients to be shared while ensuring privacy, we maintain an accuracy rate exceeding 90%. This initiative marks a significant step towards a scalable, collaborative, efficient, and dependable computational framework that aids clinical pathologists in their daily diagnostic tasks.

Optimizing community science contributions in ecology: A case study on Zooniverse's ‘Chicago wildlife watch’

Public participation in research, or community science (CS), has an important role in advancing ecological research, especially data processing. CS contributions to camera trap studies have supported wildlife conservation through the rapid processing of images and videos. However, more studies are needed to quantify the accuracy and efficiency of CS participation. We used a case study from Chicago Wildlife Watch, a Zooniverse project, to explore variability in image classification accuracy and assess efficiency of responsive retirement rules which dictate how many times an image is viewed and annotated. We found that CS participants were highly accurate when classifying empty (96.0 %) and commonly photographed species in our study area (60.14 % across all species). User agreement on species images most impacted classification accuracy, though accuracy was higher for those containing larger species and those annotated by more engaged participants. With respect to efficiency, we found that three consecutive ‘empty’ classifications from participants led to over 95 % classification accuracy in empty images and if 7 participants agreed on a species present in an image, they were accurate 98 % of the time, on average. These results further support the value of CS in ecological research and the value of applying unique project designs which consider occurrence of regional species and field systems (e.g. camera placement or ecosystem). Given these results, we encourage scientists to continue applying quantitative techniques to custom design projects to effectively use CS participants' time and maximize data accuracy.

The classification of eye diseases from fundus images based on CNN and pretrained models

Visual impairment affects more than a billion people worldwide due to insufficient care or inadequate vision screening. Computer-aided diagnosis using deep neural networks is a promising approach, it can analyse and process retinal fundus images, providing valuable reference data for doctors in clinical diagnosis or screening. This study aims to achieve an accurate classification of fundus images, including images of healthy patients as well as those with diabetic retinopathy, cataracts, and glaucoma, using a convolutional neural network (CNN) architecture and several pretrained models (AlexNet, GoogleNet, ResNet18, ResNet50, YOLOv3, and VGG 19). To enhance the training process, a mirror effect technique was applied to augment the volume of data. The experimental study resulted in very satisfactory outcomes, with the GoogleNet model paired with the SGDM optimiser achieving the highest accuracy (92.7 %).

Image classification-based transfer learning framework for image detection of IoT devices

For artificial intelligence (AI) applications, centralized learning on a cloud server and local learning on an internet-of-things device may suffer from data privacy leakage due to data sharing and inaccurate prediction due to limited computing resources. Transfer learning has been proposed as one potential solution to the world's big data problems. Transfer learning eliminates the need for each internet-of-things device to share local data with the cloud server during the training process. Instead, it can go through the training process on its own, using a cloud server's pre-trained model with high accuracy. As a result, despite its limited computing resources, the internet of things device can still predict with high accuracy. This paper proposes a transfer learning model for improving image detection accuracy on IoT devices with restricted computation. To obtain accurate image classification, a deep learning approach based on convolutional neural networks is used. The proposed method with freeze and unfreeze approaches achieves a higher validation accuracy (up to 43.6%) and a lower validation loss (up to 6.5 times) than the non-transfer learning method, according to simulation results using three relevant internet-of-things datasets.

Quantum machine learning for image classification

Image classification, a pivotal task in multiple industries, faces computational challenges due to the burgeoning volume of visual data. This research addresses these challenges by introducing two quantum machine learning models that leverage the principles of quantum mechanics for effective computations. Our first model, a hybrid quantum neural network with parallel quantum circuits, enables the execution of computations even in the noisy intermediate-scale quantum era, where circuits with a large number of qubits are currently infeasible. This model demonstrated a record-breaking classification accuracy of 99.21% on the full MNIST dataset, surpassing the performance of known quantum–classical models, while having eight times fewer parameters than its classical counterpart. Also, the results of testing this hybrid model on a Medical MNIST (classification accuracy over 99%), and on CIFAR-10 (classification accuracy over 82%), can serve as evidence of the generalizability of the model and highlights the efficiency of quantum layers in distinguishing common features of input data. Our second model introduces a hybrid quantum neural network with a Quanvolutional layer, reducing image resolution via a convolution process. The model matches the performance of its classical counterpart, having four times fewer trainable parameters, and outperforms a classical model with equal weight parameters. These models represent advancements in quantum machine learning research and illuminate the path towards more accurate image classification systems.

Can using a pre-trained deep learning model as the feature extractor in the bag-of-deep-visual-words model always improve image classification accuracy?

This article investigates whether higher classification accuracy can always be achieved by utilizing a pre-trained deep learning model as the feature extractor in the Bag-of-Deep-Visual-Words (BoDVW) classification model, as opposed to directly using the new classification layer of the pre-trained model for classification. Considering the multiple factors related to the feature extractor -such as model architecture, fine-tuning strategy, number of training samples, feature extraction method, and feature encoding method-we investigate these factors through experiments and then provide detailed answers to the question. In our experiments, we use five feature encoding methods: hard-voting, soft-voting, locally constrained linear coding, super vector coding, and fisher vector (FV). We also employ two popular feature extraction methods: one (denoted as Ext-DFs(CP)) uses a convolutional or non-global pooling layer, and another (denoted as Ext-DFs(FC)) uses a fully-connected or global pooling layer. Three pre-trained models-VGGNet-16, ResNext-50(32×4d), and Swin-B-are utilized as feature extractors. Experimental results on six datasets (15-Scenes, TF-Flowers, MIT Indoor-67, COVID-19 CXR, NWPU-RESISC45, and Caltech-101) reveal that compared to using the pre-trained model with only the new classification layer re-trained for classification, employing it as the feature extractor in the BoDVW model improves the accuracy in 35 out of 36 experiments when using FV. With Ext-DFs(CP), the accuracy increases by 0.13% to 8.43% (averaged at 3.11%), and with Ext-DFs(FC), it increases by 1.06% to 14.63% (averaged at 5.66%). Furthermore, when all layers of the pre-trained model are fine-tuned and used as the feature extractor, the results vary depending on the methods used. If FV and Ext-DFs(FC) are used, the accuracy increases by 0.21% to 5.65% (averaged at 1.58%) in 14 out of 18 experiments. Our results suggest that while using a pre-trained deep learning model as the feature extractor does not always improve classification accuracy, it holds great potential as an accuracy improvement technique.

Self-supervised CondenseNet for feature learning to increase the accuracy in image classification

Self-supervised CondenseNet for feature learning to increase the accuracy in image classification

TIM‐Net: A multi‐label classification network for TCM tongue images fusing global‐local features

AbstractCombining the extracted tongue features with other medical indicators can effectively judge the diseases of patients. The previous work usually only analyzes a certain feature of the tongue body and is unable to extract multiple features simultaneously. In this study, a multi‐label classification network named TIM‐Net is proposed, which integrates global and local features to achieve multi‐label intelligent diagnosis of Chinese medicine tongue images. First, a feature extraction network based on ResNet is proposed to capture the features of tongue images more sufficiently. Then, a multi‐label classification algorithm fusing global and local features is proposed, and targeted screening operations are carried out on the class‐related feature maps based on global confidence. In addition, a logical masking algorithm is proposed to ensure that the local features can only correct the feature labels they represent, and do not interfere with other feature labels. The classification accuracy is further improved by using local feature confidence and correcting the global classification results. Finally, the experimental results indicate that the classification accuracy of the tongue images is gradually improved through optimizing the feature extraction network and fusing local features, and it exceeds other state‐of‐the‐art multi‐label classification networks.

Image Segmentation and Classification Fields Are Generated Using Binary Random Fields Based on Planar Graph and Neighborhood Spanning Tree

Image analysis plays a pivotal role in computer vision, with image segmentation and classification being fundamental tasks in this domain. This abstract presents a novel approach to image processing that leverages Binary Random Fields (BRF) with a foundation in planar graphs and neighborhood spanning trees. This innovative methodology seeks to enhance the accuracy and efficiency of image segmentation and classification, addressing key challenges in computer vision applications. Binary Random Fields (BRF) is probabilistic graphical models that have proven effective in capturing spatial dependencies and contextual information within images. Our proposed method extends the utility of BRF by incorporating planar graph theory and neighborhood spanning trees to refine the segmentation and classification processes. Planar graphs offer a structured representation of image data, preserving topological relationships among pixels, while neighborhood spanning trees provide a hierarchical framework for modeling image regions

A novel automatic annotation method for whole slide pathological images combined clustering and edge detection technique

AbstractPixel‐level labeling of regions of interest in an image is a key step in building a labeled training dataset for supervised deep learning networks of images. However, traditional manual labeling of cancerous regions in digital pathological images by doctors is time‐consuming and inefficient. To address this issue, this paper proposes an automatic labeling method for whole slide images, which combines clustering and edge detection techniques. The proposed method utilizes the multi‐level feature fusion model and the Long‐Short Term Memory network to discriminate the cancerous nature of the whole slide images, thereby improving the classification accuracy of the whole slide images. Subsequently, the automatic labeling of cancerous regions is achieved by integrating a density‐based clustering algorithm and an edge point extraction algorithm, both based on the discriminated results of the cancerous properties of whole slide images. The experimental results demonstrate the effectiveness of the proposed method, which offers an efficient and accurate solution to the challenging task of cancerous region labeling in digital pathological images.

Adapting transfer learning models to dataset through pruning and Avg-TopK pooling

This study focuses on efficiently adapting transfer learning models to address the challenges of creating customized deep learning models for specific datasets. Designing a model from scratch can be time-consuming and complex due to factors like model complexity, size, and dataset structure. To overcome these obstacles, a novel approach is proposed using transfer learning models. The proposed method involves identifying relevant layers in transfer learning models and removing unnecessary ones using a layer-based variance pruning technique. This results in the creation of new models with improved computational efficiency and classification performance. By streamlining the models through layer-based variance pruning, the study achieves enhanced accuracy and faster computation. Experiments were conducted using the COVID-19 dataset and well-known transfer learning models, including InceptionV3, ResNet50V2, DenseNet201, VGG16, and Xception to validate the approach. Among these models, the variance-based layer pruning technique was applied to InceptionV3 and DenseNet201, yielding the best results. When these pruned models were combined with the new pooling layer, Avg-TopK, the proposed method achieved an outstanding image classification accuracy of 99.3%. Comparisons with previous models and literature studies indicate that the proposed approach outperforms existing methods, showcasing state-of-the-art performance. This high-performance approach provides great potential for diagnosing COVID-19 and monitoring disease progression, especially on hardware-limited devices. By leveraging transfer learning models, pruning, and efficient pooling techniques, the study presents a promising strategy for tackling challenges in custom model design, leading to exceptional results in such as image classification and segmentation tasks. The proposed methodology holds the potential to yield exceptional outcomes across a spectrum of tasks, encompassing disciplines such as image classification and segmentation.

An Effective Randomized Algorithm for Hyperspectral Image Feature Extraction

Analyzing the spectral and spatial characteristics of Hyperspectral Imaging (HSI) in a three-dimensional space is a challenging task. Recently, there have been developments in 3D feature extraction methods based on tensor decomposition, which allow for the effective utilization of both global and local information in HSI. These methods also explore the inherent low-rank properties of HSI through tensor decomposition. In this paper, we propose a new approach called variable randomized T-product decomposition (Vrt-SVD), which is a variation of Tensor Singular Spectral Analysis. The goal of this approach is to improve the efficiency of tensor methods for feature extraction and reduce artifacts of image processing. By using a randomized algorithm based on the variable t-SVD, we are able to capture both global and local spatial and spectral information in HSI efficiently, which enables us to explore its low-rank characteristics. To evaluate the effectiveness of the extracted features, we use a Support Vector Machine (SVM) classifier to assess the accuracy of image classification. By conducting numerous numerical experiments, we provide strong evidence to show that the proposed method outperforms several advanced feature extraction techniques.

Feature decoupling and interaction network for defending against adversarial examples

Recently, it was found that deep neural networks (DNNs) are susceptible to adversarial input perturbations. Most defense strategies adopt the denoising method based on preprocessing, which mitigates the impacts of adversarial perturbations on DNNs by learning the distributions of nonadversarial datasets and projecting adversarial inputs into the learned nonadversarial manifolds. However, existing defense strategies commonly focus on reconstructing clean images while ignoring the role of adversarial perturbations, which results in the reconstructed images failing to achieve the visual quality and classification accuracy of the original clean images, and the induced adversarial robustness improvement is limited. This paper proposes a feature decoupling-interaction network (FDIN), which introduces the concepts of clean features and adversarial features to separate the two kinds of features from the input adversarial examples (AEs) in a feature decoupling-interaction manner. The clean features are used to reconstruct the input image so that it is infinitely close to the original clean image, and the adversarial features are used to reconstruct the adversarial perturbations. Adversarial perturbations are removed from the adversarial examples across multiple cross cycles to improve further the reconstructed image's visual quality and classification accuracy. The features of the original clean image are used as prior knowledge to guide the network to learn the clean features of the adversarial examples and improve the classification accuracy of the model on the clean examples. In addition, a classification loss function based on the Carlini & Wagner (CW) attack algorithm is used instead of the conventional cross-entropy loss function to improve the adversarial robustness of the FDIN. The experimental results show that the proposed method achieves better defense performance than the current state-of-the-art methods on both standard tests and various attack tests and even exceeds the test accuracy of the target classifier on the original test set.

Enhancing skin lesion segmentation with U-Net++: Design, analysis, and performance evaluation

The present research examines the enhancement of skin lesion segmentation with U-Net++. Achieving accurate classification of dermoscopy images is heavily contingent on the precise segmentation of skin lesions or nodules. However, this task is considerably challenging due to the elusive edges, irregular perimeters, and variations both within and across lesion classes. Despite numerous existing algorithms for segmenting skin lesions from dermoscopic images, they often fall short of industry benchmarks in terms of precision. To address this, our research introduces a novel U-Net++ architecture, implementing tailored adjustments to feature map dimensions. The aim is to significantly enhance automated segmentation precision for dermoscopic images. Our evaluation involved a comprehensive assessment of the model's performance, encompassing an exploration of various parameters such as epochs, batch size, and optimizer selections. Additionally, we conducted extensive testing using augmentation techniques to bolster the image volume within the HAM10000 dataset. A key innovation in our research is the integration of a hair removal process into the U-Net++ algorithm, significantly enhancing image quality and subsequently leading to improved segmentation accuracy. The results of our proposed method demonstrate substantial advancements, showcasing an impressive Mean Intersection over Union (IoU) of 84.1%, a Mean Dice Coefficient of 91.02%, and a Segmentation Test Accuracy of 95.10%. Our suggested U-Net++ algorithm does a better job of segmenting than U-Net, Modified U-Net, K-Nearest Neighbors (KNN), and Support Vector Machine (SVM). This shows that it could be used to improve dermoscopy image analysis. Our proposed algorithm shows remarkable improvement in both observational outcomes and classifier performance.