Learning For Image Classification Research Articles

Enhancing deep learning image classification using data augmentation and genetic algorithm-based optimization

Enhancing deep learning image classification using data augmentation and genetic algorithm-based optimization

Comparing hybrid models for recognising objects in thermal images at nighttime

This research aims to revolutionize urban object recognition by developing cloud-based Python programs using intelligent algorithms. Unlike current models that focus on colour enhancement in nighttime thermal images, this work addresses the critical challenge of accurate object detection in urban landscapes. The proposed method incorporates a binary generative adversarial network (GAN) generator that can switch bidirectionally between daytime colour (DC) and nighttime infrared (NTIR) images. memory-based visual image memory (MVAM), system extracts important descriptive information from urban landscape images, reducing problems related to small sample sizes. This discussion presents a comprehensive improvement and evaluation of a deep learning image classification pipeline using Google Colab, demonstrating advanced image processing. Using TensorFlow, Keres and scikit image libraries combined with advanced algorithms such as DenseNet121 and MobileNetV2 presents a clear approach. We created a Bidirectional GAN + MVAM for object recognition in this work. Our method performed well, with an accuracy of 81.43%, precision of 51.16, recall of 50.11, and F-score of 46.37. The systematic presentation of the code presents a careful strategy to ensure optimal performance, stability, and efficiency of deep learning and image processing tasks.

The Role of ArtificiaI Intelligence in Brain Tumor Diagnosis: An Evaluation of a Machine Learning Model.

This research study explores of the effectiveness of a machine learning image classification model in the accurate identification of various types of brain tumors. The types of tumors under consideration in this study are gliomas, meningiomas, and pituitary tumors. These are some of the most common types of brain tumors and pose significant challenges in terms of accurate diagnosis and treatment. The machine learning model that is the focus of this study is built on the Google Teachable Machine platform (Alphabet Inc., Mountain View, CA). The Google Teachable Machine is a machine learning image classification platform that is built from Tensorflow, a popular open-source platform for machine learning. The Google Teachable Machine model was specifically evaluated for its ability to differentiate between normal brains and the aforementioned types of tumors in MRI images. MRI images are a common tool in the diagnosis of brain tumors, but the challenge lies in the accurate classification of the tumors. This is where the machine learning model comes into play. The model is trained to recognize patterns in the MRI images that correspond to the different types of tumors. The performance of the machine learning model was assessed using several metrics. These include precision, recall, and F1 score. These metrics were generated from a confusion matrix analysis and performance graphs. A confusion matrix is a table that is often used to describe the performance of a classification model. Precision is a measure of the model's ability to correctly identify positive instances among all instances it identified as positive. Recall, on the other hand, measures the model's ability to correctly identify positive instances among all actual positive instances. The F1 score is a measure that combines precision and recall providing a single metric for model performance. The results of the study were promising. The Google Teachable Machine model demonstrated high performance, with accuracy, precision, recall, and F1 scores ranging between 0.84 and 1.00. This suggests that the model is highly effective in accurately classifying the different types of brain tumors. This study provides insights into the potential of machine learning models in the accurate classification of brain tumors. The findings of this study lay the groundwork for further research in this area and have implications for the diagnosis and treatment of brain tumors. The study also highlights the potential of machine learning in enhancing the field of medical imaging and diagnosis. With the increasing complexity and volume of medical data, machine learning models like the one evaluated in this study could play a crucial role in improving the accuracy and efficiency of diagnoses. Furthermore, the study underscores the importance of continued research and development in this field to further refine these models and overcome any potential limitations or challenges. Overall, the study contributes to the field of medical imaging and machine learning and sets the stage for future research and advancements in this area.

Colour fusion effect on deep learning classification of uveal melanoma.

Reliable differentiation of uveal melanoma and choroidal nevi is crucial to guide appropriate treatment, preventing unnecessary procedures for benign lesions and ensuring timely treatment for potentially malignant cases. The purpose of this study is to validate deep learning classification of uveal melanoma and choroidal nevi, and to evaluate the effect of colour fusion options on the classification performance. A total of 798 ultra-widefield retinal images of 438 patients were included in this retrospective study, comprising 157 patients diagnosed with UM and 281 patients diagnosed with choroidal naevus. Colour fusion options, including early fusion, intermediate fusion and late fusion, were tested for deep learning image classification with a convolutional neural network (CNN). F1-score, accuracy and the area under the curve (AUC) of a receiver operating characteristic (ROC) were used to evaluate the classification performance. Colour fusion options were observed to affect the deep learning performance significantly. For single-colour learning, the red colour image was observed to have superior performance compared to green and blue channels. For multi-colour learning, the intermediate fusion is better than early and late fusion options. Deep learning is a promising approach for automated classification of uveal melanoma and choroidal nevi. Colour fusion options can significantly affect the classification performance.

Deep learning and random light structuring ensure robust free-space communications

Having shown early promise, free-space optical (FSO) communications face formidable challenges in the age of information explosion. The ever-growing demand for greater channel communication capacity is one of the challenges. The inter-channel crosstalk, which severely degrades the quality of transmitted information, creates another roadblock in the way of efficient implementation of FSO communication systems. Here, we advance theoretically and realize experimentally a potentially high-capacity FSO protocol that enables high-fidelity transfer of an image or set of images through a complex environment. In our protocol, we complement random light structuring at the transmitter with a deep learning image classification platform at the receiver. Multiplexing unique, independent, mutually orthogonal degrees of freedom available to structured random light can potentially significantly boost the channel communication capacity of our protocol without introducing any deleterious crosstalk. Specifically, we show how one can multiplex the degrees of freedom associated with the source coherence radius and a spatial position of a beamlet within an array of structured random beams to greatly enhance the capacity of our communication link. The superb resilience of structured random light to environmental noise, as well as extreme efficiency of deep learning networks at classifying images, guarantees high-fidelity image transfer within the framework of our protocol.

Systematic review and best practices for drone remote sensing of invasive plants

Abstract Drones have emerged as a cost‐effective solution to detect and map plant invasions, offering researchers and land managers flexibility in flight design, sensors and data collection schedules. A systematic review of trends in drone‐based image collection, data processing and analytical approaches is needed to advance the science of invasive species monitoring and management and improve scalability and replicability. We systematically reviewed studies using drones for plant invasion research to identify knowledge gaps, best practices and a path toward advancing the science of invasive plant monitoring and management. We devised a database of 33 standardized reporting parameters, coded each study to those parameters, calculated descriptive statistics and synthesized how these technologies are being implemented and used. Trends show a general increase in studies since 2009 with a bias toward temperate regions in North America and Europe. Most studies have focused on testing the validity of a machine learning or deep learning image classification technique with fewer studies focused on monitoring or modelling spread. Very few studies used drones for assessing ecosystem dynamics and impacts such as determining environmental drivers or tracking re‐emergence after disturbance. Overall, we noted a lack of standardized reporting on field survey design, flight design, drone systems, image processing and analyses, which hinders replicability and scalability of approaches. Based on these findings, we develop a standard framework for drone applications in invasive species monitoring to foster cross‐study comparability and reproducibility. We suggest several areas for advancing the use of drones in invasive plant studies including (1) utilizing standardized reporting frameworks to facilitate scientific research practices, (2) integrating drone data with satellite imagery to scale up relationships over larger areas, (3) using drones as an alternative to in‐person ground surveys and (4) leveraging drones to assess community trait shifts tied to plant fitness and reproduction.

Recommendation System Building based on CNN and TF-IDF Approaches

Online shopping has becoming increasingly popular. Customers remain engaged with shopping sites due to effective product recommendations. By analyzing users’ past actions and preferences, shopping platforms effectively identify customer interests and desires, enhancing their overall shopping experience. This paper presents a recommendation system that displays 5 items relevant to the product currently being viewed by the customer. The system processes images with an adapted ResNet-50, a deep learning image classification model in convolutional neural network (CNN) and use the acquired embedded vector to determine similarities with cosine functions. Additionally, the system employs term frequency-inverse document frequency (TF-IDF) method for text analysis in product descriptions, generating word embedding that assists recommendations. This blend of visual and textual analysis ensures that the suggestions closely match the item’s category. The system achieves 94.72% accuracy in subcategories’ classification, a 4% increase compared to using the CNN method alone, confirming its effectiveness in recommending relevant items.

A novel women's ovulation prediction through salivary ferning using the box counting and deep learning

There are several methods to predict a woman's ovulation time, including using a calendar system, basal body temperature, ovulation prediction kit, and OvuScope. This is the first study to predict the time of ovulation in women by calculating the results of detecting the fractal shape of the full ferning (FF) line pattern in salivary using pixel counting, box counting, and deep learning for computer vision methods. The peak of a woman's ovulation every month in her menstrual cycle occurs when the number of ferning lines is the most numerous or dense, and this condition is called FF. In this study, the computational results based on the visualization of the fractal shape of the salivary ferning line pattern from the pixel-counting method have an accuracy of 80%, while the fractal dimensions achieved by the box-counting are 1.474. On the other hand, using the deep learning image classification, we obtain the highest accuracy of 100% with a precision value of 1.00, recall of 1.00, and F1-score 1.00 on the pre-trained network model ResNet-18. Furthermore, visualization of the ResNet-34 model results in the highest number of patches, i.e., 586 patches (equal to 36,352 pixels), by applying fern-like lines pattern detection with windows size 8x8 pixels.

An improved post-hurricane building damaged detection method based on transfer learning

After a natural disaster, it is very important for the government to conduct a damaged assessment as soon as possible. Fast and accurate disaster assessment helps the government disaster relief departments allocate resources and respond quickly and effectively to minimize the losses caused by the disaster. Usually, the method of measuring disaster losses is to rely on manual field exploration and measurement, and then calculate and label the damaged buildings or land, or rely on unmanned collections to remotely collect pictures of the disaster-stricken area, and compare the original pictures to carry out the disaster annotation and calculation. These methods are time-consuming, labor-intensive, and inefficient. This paper proposes a post-hurricane building damage detection method based on transfer learning, which uses deep learning image classification algorithms to achieve post-disaster satellite image damage detection and classification, thereby improving disaster assessment efficiency and preparing for disaster relief and post-disaster reconstruction. The proposed method adopts the theory of transfer learning, establishes a disaster image detection model based on the convolutional neural network model, and uses the 2017 Hurricane Harvey data as the experimental data set. Experiments have proved that our proposed model accuracy of disaster detection reaches 97%, which is 1% higher than other models.

Alterations in zonal distribution and plasma membrane localization of hepatocyte bile acid transporters in patients with NAFLD.

NAFLD is highly prevalent with limited treatment options. Bile acids (BAs) increase in the systemic circulation and liver during NAFLD progression. Changes in plasma membrane localization and zonal distribution of BA transporters can influence transport function and BA homeostasis. However, a thorough characterization of how NAFLD influences these factors is currently lacking. This study aimed to evaluate the impact of NAFLD and the accompanying histologic features on the functional capacity of key hepatocyte BA transporters across zonal regions in human liver biopsies. A novel machine learning image classification approach was used to quantify relative zonal abundance and plasma membrane localization of BA transporters (bile salt export pump [BSEP], sodium-taurocholate cotransporting polypeptide, organic anion transporting polypeptide [OATP] 1B1 and OATP1B3) in non-diseased (n = 10), NAFL (n = 9), and NASH (n = 11) liver biopsies. Based on these data, membrane-localized zonal abundance (MZA) measures were developed to estimate transporter functional capacity. NAFLD diagnosis and histologic scoring were associated with changes in transporter membrane localization and zonation. Increased periportal BSEPMZA (mean proportional difference compared to non-diseased liver of 0.090) and decreased pericentral BSEPMZA (-0.065) were observed with NASH and also in biopsies with higher histologic scores. Compared to Non-diseased Liver, periportal OATP1B3MZA was increased in NAFL (0.041) and NASH (0.047). Grade 2 steatosis (mean proportional difference of 0.043 when compared to grade 0) and grade 1 lobular inflammation (0.043) were associated with increased periportal OATP1B3MZA. These findings provide novel mechanistic insight into specific transporter alterations that impact BA homeostasis in NAFLD. Changes in BSEPMZA likely contribute to altered BA disposition and pericentral microcholestasis previously reported in some patients with NAFLD. BSEPMZA assessment could inform future development and optimization of NASH-related pharmacotherapies.

Unlabeled data selection for active learning in image classification

Active Learning has emerged as a viable solution for addressing the challenge of labeling extensive amounts of data in data-intensive applications such as computer vision and neural machine translation. The main objective of Active Learning is to automatically identify a subset of unlabeled data samples for annotation. This identification process is based on an acquisition function that assesses the value of each sample for model training. In the context of computer vision, image classification is a crucial task that typically requires a substantial training dataset. This research paper introduces innovative selection methods within the Active Learning framework, aiming to identify informative images from unlabeled datasets while minimizing the number of required training data. The proposed methods, namely Similari-ty-based Selection, Prediction Probability-based Selection, and Competence-based Active Learning, have been extensively evaluated through experiments conducted on popular datasets like Cifar10 and Cifar100. The experimental results demonstrate that the proposed methods outperform random selection and conventional selection techniques. The superior performance of the novel selection methods underscores their effectiveness in enhancing the Active Learning process for image classification tasks.

Application of Semi-Supervised Learning in Image Classification: Research on Fusion of Labeled and Unlabeled Data

Application of Semi-Supervised Learning in Image Classification: Research on Fusion of Labeled and Unlabeled Data

Machine learning classification based on k-Nearest Neighbors for PolSAR data.

In this work, we focus on obtaining insights of the performances of some well-known machine learning image classification techniques (k-NN, Support Vector Machine, randomized decision tree and one based on stochastic distances) for PolSAR (Polarimetric Synthetic Aperture Radar) imagery. We test the classifiers methods on a set of actual PolSAR data and provide some conclusions. The aim of this work is to show that suitable adapted standard machine learning methods offer excellent performances vs.computational complexity trade-off for PolSAR image classification. In this work, we evaluate well-known machine learning techniques for PolSAR (Polarimetric Synthetic Aperture Radar) image classification, including K-Nearest Neighbors (KNN), Support Vector Machine (SVM), randomized decision tree, and a method based on the Kullback-Leibler stochastic distance. Our experiments with real PolSAR data show that standard machine learning methods, when adapted appropriately, offer a favourable trade-off between performance and computational complexity. The KNN and SVM perform poorly on these data, likely due to their failure to account for the inherent speckle presence and properties of the studied reliefs. Overall, our findings highlight the potential of the Kullback-Leibler stochastic distance method for PolSAR image classification.

Improving Performance of Impact Point Determination Algorithms of Gamma Interaction in a Monolithic Scintillator Using ML and Numerical Methods

Two algorithms for classification of the number of interaction points of gamma-ray in a Compton camera with scintillator crystal are developed, taking into account the limited processing power of the edge electronics in the space environment. One of the algorithms is relaying on a machine learning image classification convolution network, and the other is a numerical method, searching for multiple maxima peaks in the light intensity over the detector plane. Preliminary results of the accuracy of the two approaches are provided and discussed.

An Approach for Egg Parasite Classification Based on Ensemble Deep Learning

Opisthorchis viverrini and minute intestinal fluke (MIF) infections are heavily epidemic in northeastern Thailand. Their primary cause is eating raw or undercooked cyprinid fishes, and they cause health problems in the human digestive system. In cases of liver fluke, these parasites can go through the bile duct system, which may cause cholangiocarcinoma (bile duct cancer). When a medical doctor suspects that a patient is infected with parasites, they typically request a stool analysis to determine the type of egg parasites using microscopy. Both parasites have similar characteristics, thus, it is necessary for a specialist to identify the specific type of egg parasites present. Many automatic systems have been developed using deep learning to assist doctors in diagnosing the type of egg parasite. In this study, we proposed three models of deep learning architectures and created voting ensembles to analyze egg parasite images. Images of similar liver fluke eggs and MIF eggs were taken from the Parasitology Laboratory, Faculty of Medicine, Khon Kaen University, Khon Kaen, Thailand. Image data augmentation is used to expand images from different perspectives and assist the system in acquiring a greater variety of images. Three models performed effectively, by employing the hard voting ensemble, the accuracy increased to 86.67%, while for the second group, the accuracies reached 68.00%, 76.00%, and 77.33%, respectively. Using the soft voting ensemble, the accuracy improved to 79.33%. These outcomes highlight the potential of ensemble deep learning in image classification. Furthermore, these results align closely with those achieved by several experts in image classification. Hence, a promising ensemble approach can aid doctors in accurately classifying images of egg parasites.

An extended focused assessment with sonography in trauma ultrasound tissue-mimicking phantom for developing automated diagnostic technologies.

Introduction: Medical imaging-based triage is critical for ensuring medical treatment is timely and prioritized. However, without proper image collection and interpretation, triage decisions can be hard to make. While automation approaches can enhance these triage applications, tissue phantoms must be developed to train and mature these novel technologies. Here, we have developed a tissue phantom modeling the ultrasound views imaged during the enhanced focused assessment with sonography in trauma exam (eFAST). Methods: The tissue phantom utilized synthetic clear ballistic gel with carveouts in the abdomen and rib cage corresponding to the various eFAST scan points. Various approaches were taken to simulate proper physiology without injuries present or to mimic pneumothorax, hemothorax, or abdominal hemorrhage at multiple locations in the torso. Multiple ultrasound imaging systems were used to acquire ultrasound scans with or without injury present and were used to train deep learning image classification predictive models. Results: Performance of the artificial intelligent (AI) models trained in this study achieved over 97% accuracy for each eFAST scan site. We used a previously trained AI model for pneumothorax which achieved 74% accuracy in blind predictions for images collected with the novel eFAST tissue phantom. Grad-CAM heat map overlays for the predictions identified that the AI models were tracking the area of interest for each scan point in the tissue phantom. Discussion: Overall, the eFAST tissue phantom ultrasound scans resembled human images and were successful in training AI models. Tissue phantoms are critical first steps in troubleshooting and developing medical imaging automation technologies for this application that can accelerate the widespread use of ultrasound imaging for emergency triage.

Improved deep learning image classification algorithm based on Swin Transformer V2

While convolutional operation effectively extracts local features, their limited receptive fields make it challenging to capture global dependencies. Transformer, on the other hand, excels at global modeling and effectively captures global dependencies. However, the self-attention mechanism used in Transformers lacks a local mechanism for information exchange within specific regions. This article attempts to leverage the strengths of both Transformers and convolutional neural networks (CNNs) to enhance the Swin Transformer V2 model. By incorporating both convolutional operation and self-attention mechanism, the enhanced model combines the local information-capturing capability of CNNs and the long-range dependency-capturing ability of Transformers. The improved model enhances the extraction of local information through the introduction of the Swin Transformer Stem, inverted residual feed-forward network, and Dual-Branch Downsampling structure. Subsequently, it models global dependencies using the improved self-attention mechanism. Additionally, downsampling is applied to the attention mechanism’s Q and K to reduce computational and memory overhead. Under identical training conditions, the proposed method significantly improves classification accuracy on multiple image classification datasets, showcasing more robust generalization capabilities.

Omni-Modeler: Rapid Adaptive Visual Recognition with Dynamic Learning

Deep neural network (DNN) image classification has grown rapidly as a general pattern detection tool for an extremely diverse set of applications; yet dataset accessibility remains a major limiting factor for many applications. This paper presents a novel dynamic learning approach to leverage pretrained knowledge to novel image spaces in the effort to extend the algorithm knowledge domain and reduce dataset collection requirements. The proposed Omni-Modeler generates a dynamic knowledge set by reshaping known concepts to create dynamic representation models of unknown concepts. The Omni-Modeler embeds images with a pretrained DNN and formulates compressed language encoder. The language encoded feature space is then used to rapidly generate a dynamic dictionary of concept appearance models. The results of this study demonstrate the Omni-Modeler capability to rapidly adapt across a range of image types enabling the usage of dynamically learning image classification with limited data availability.

Comparative Analysis of Machine Learning Models for Image Detection of Colonic Polyps vs. Resected Polyps.

(1) Background: Colon polyps are common protrusions in the colon's lumen, with potential risks of developing colorectal cancer. Early detection and intervention of these polyps are vital for reducing colorectal cancer incidence and mortality rates. This research aims to evaluate and compare the performance of three machine learning image classification models' performance in detecting and classifying colon polyps. (2) Methods: The performance of three machine learning image classification models, Google Teachable Machine (GTM), Roboflow3 (RF3), and You Only Look Once version 8 (YOLOv8n), in the detection and classification of colon polyps was evaluated using the testing split for each model. The external validity of the test was analyzed using 90 images that were not used to test, train, or validate the model. The study used a dataset of colonoscopy images of normal colon, polyps, and resected polyps. The study assessed the models' ability to correctly classify the images into their respective classes using precision, recall, and F1 score generated from confusion matrix analysis and performance graphs. (3) Results: All three models successfully distinguished between normal colon, polyps, and resected polyps in colonoscopy images. GTM achieved the highest accuracies: 0.99, with consistent precision, recall, and F1 scores of 1.00 for the 'normal' class, 0.97-1.00 for 'polyps', and 0.97-1.00 for 'resected polyps'. While GTM exclusively classified images into these three categories, both YOLOv8n and RF3 were able to detect and specify the location of normal colonic tissue, polyps, and resected polyps, with YOLOv8n and RF3 achieving overall accuracies of 0.84 and 0.87, respectively. (4) Conclusions: Machine learning, particularly models like GTM, shows promising results in ensuring comprehensive detection of polyps during colonoscopies.

Rolling Bearing Fault Diagnosis Based on SVD-GST Combined with Vision Transformer

Aiming at rolling bearing fault diagnosis, the collected vibration signal contains complex noise interference, and one-dimensional information cannot be used to fully mine the data features of the problem. This paper proposes a rolling bearing fault diagnosis method based on SVD-GST combined with the Vision Transformer. Firstly, the one-dimensional vibration signal is preprocessed to reduce noise using singular value decomposition (SVD) to obtain a more accurate and useful signal. Then, the generalized S-transform (GST) is used to convert the processed one-dimensional vibration signal into a two-dimensional time–frequency image and make full use of the advantages of deep learning in image classification with higher recognition accuracy. In order to avoid the problem of limited sensory fields in CNN and the need for an RNN to compute step by step over time when processing sequence data, the use of a Vision Transformer model for pattern recognition classification is proposed. Finally, an experimental platform for the fault diagnosis of rolling bearings is built. The model is experimentally validated, achieving an average accuracy of 98.52% over multiple tests. Additionally, compared with the SVD-GST-2DCNN, STFT-CNN-LSTM, SVD-GST-LSTM, and GST-ViT fault diagnosis models, the proposed method has higher diagnostic accuracy and stability, providing a new method for rolling bearing fault diagnosis.