Remote Classification Research Articles

Secure and Transparent Lung and Colon Cancer Classification Using Blockchain and Microsoft Azure

Background: The global healthcare system faces challenges in diagnosing and managing lung and colon cancers, which are significant health burdens. Traditional diagnostic methods are inefficient and prone to errors, while data privacy and security concerns persist. Objective: This study aims to develop a secure and transparent framework for remote consultation and classification of lung and colon cancer, leveraging blockchain technology and Microsoft Azure cloud services. Dataset and Features: The framework utilizes the LC25000 dataset, containing 25,000 histopathological images, for training and evaluating advanced machine learning models. Key features include secure data upload, anonymization, encryption, and controlled access via blockchain and Azure services. Methods: The proposed framework integrates Microsoft Azure’s cloud services with a permissioned blockchain network. Patients upload CT scans through a mobile app, which are then preprocessed, anonymized, and stored securely in Azure Blob Storage. Blockchain smart contracts manage data access, ensuring only authorized specialists can retrieve and analyze the scans. Azure Machine Learning is used to train and deploy state-of-the-art machine learning models for cancer classification. Evaluation Metrics: The framework’s performance is evaluated using metrics such as accuracy, precision, recall, and F1-score, demonstrating the effectiveness of the integrated approach in enhancing diagnostic accuracy and data security. Results: The proposed framework achieves an impressive accuracy of 100% for lung and colon cancer classification using DenseNet, ResNet50, and MobileNet models with different split ratios (70–30, 80–20, 90–10). The F1-score and k-fold cross-validation accuracy (5-fold and 10-fold) also demonstrate exceptional performance, with values exceeding 99.9%. Real-time notifications and secure remote consultations enhance the efficiency and transparency of the diagnostic process, contributing to better patient outcomes and streamlined cancer care management.

Automated classification of mixed populations of Aedes aegypti and Culex quinquefasciatus mosquitoes under field conditions

BackgroundThe recent rise in the transmission of mosquito-borne diseases such as dengue virus (DENV), Zika (ZIKV), chikungunya (CHIKV), Oropouche (OROV), and West Nile (WNV) is a major concern for public health managers worldwide. Emerging technologies for automated remote mosquito classification can be supplemented to improve surveillance systems and provide valuable information regarding mosquito vector catches in real time.MethodsWe coupled an optical sensor to the entrance of a standard mosquito suction trap (BG-Mosquitaire) to record 9151 insect flights in two Brazilian cities: Rio de Janeiro and Brasilia. The traps and sensors remained in the field for approximately 1 year. A total of 1383 mosquito flights were recorded from the target species: Aedes aegypti and Culex quinquefasciatus. Mosquito classification was based on previous models developed and trained using European populations of Aedes albopictus and Culex pipiens.ResultsThe VECTRACK sensor was able to discriminate the target mosquitoes (Aedes and Culex genera) from non-target insects with an accuracy of 99.8%. Considering only mosquito vectors, the classification between Aedes and Culex achieved an accuracy of 93.7%. The sex classification worked better for Cx. quinquefasciatus (accuracy: 95%; specificity: 95.3%) than for Ae. aegypti (accuracy: 92.1%; specificity: 88.4%).ConclusionsThe data reported herein show high accuracy, sensitivity, specificity and precision of an automated optical sensor in classifying target mosquito species, genus and sex. Similar results were obtained in two different Brazilian cities, suggesting high reliability of our findings. Surprisingly, the model developed for European populations of Ae. albopictus worked well for Brazilian Ae. aegypti populations, and the model developed and trained for Cx. pipiens was able to classify Brazilian Cx. quinquefasciatus populations. Our findings suggest this optical sensor can be integrated into mosquito surveillance methods and generate accurate automatic real-time monitoring of medically relevant mosquito species.Graphical

Adopting artificial intelligence algorithms for remote fetal heart rate monitoring and classification using wearable fetal phonocardiography

Fetal phonocardiography (FPCG) is a non-invasive Fetal Heart Rate (FHR) monitoring technique that can detect vibrations and murmurs in heart sounds. However, acquiring fetal heart sounds from a wearable FPCG device is challenging due to noise and artefacts. This research contributes a resilient solution to overcome the conventional issues by adopting Artificial Intelligence (AI) with FPCG for automated FHR monitoring in an end-to-end manner, named (AI-FHR). Four sequential methodologies were used to ensure reliable and accurate FHR monitoring. The proposed method removes low-frequency noises and high-frequency noises by using Chebyshev II high-pass filters and Enhanced Complete Ensemble Empirical Mode Decomposition with Adaptive Noise (ECEEMDAN) in combination with Phase Shifted Maximal Overlap Discrete Wavelet Transform (PS-MODWT) filters, respectively. The denoised signals are segmented to reduce complexity, and the segmentation is performed using multi-agent deep Q-learning (MA-DQL). The segmented signal is provided to reduce the redundancies in cardiac cycles using the Artificial Hummingbird Optimization (AHBO) algorithm. The segmented and non-redundant signals are converted into 3D spectrograms using a machine learning algorithm called variational auto-encoder-general adversarial networks (VAE-GAN). The feature extraction and classification are carried out by adopting a hybrid of the bidirectional gated recurrent unit (BiGRU) and the multi-boosted capsule network (MBCapsNet). The proposed method was implemented and simulated using MATLAB R2020a and validated by adopting effective validation metrics. The results demonstrate that the proposed method performed better than the current method with accuracy (81.34%), sensitivity (72%), F1-score (83%), Energy (0.808 J), and complexity index (13.34). Like other optimization methods, AHO needs precise parameter adjustment in order to function well. Its performance may be greatly impacted by the selection of parameters, including population size, exploration rate, and learning rate.

Remote 3D Imaging and Classification of Pelagic Microorganisms with A Short‐Range Multispectral Confocal LiDAR

AbstractPlankton is essential to maintain healthy aquatic ecosystems since it influences the biological carbon pump globally. However, climate change‐induced alterations to oceanic properties threaten planktonic communities. It is therefore crucial to monitor their abundance to assess the health status of marine ecosystems. In situ optical tools unlock high‐resolution measurements of sub‐millimeter specimens, but state‐of‐the‐art underwater imaging techniques are limited to fixed and small close‐range volumes, requiring the instruments to be vertically dived. Here, a novel scanning multispectral confocal light detection and ranging (LiDAR) system for short‐range volumetric sensing in aquatic media is introduced. The system expands the inelastic confocal principle to multiple wavelength channels, allowing the acquisition of 4D point clouds combining near‐diffraction limited morphological and spectroscopic data that is used to train artificial intelligence (AI) models. Volumetric mapping and classification of microplastics is demonstrated to sort them by color and size. Furthermore, in vivo autofluorescence is resolved from a community of free‐swimming zooplankton and microalgae, and accurate spectral identification of different genera is accomplished. The deployment of this photonic platform alongside AI models overcomes the complex and subjective task of manual plankton identification and enables non‐intrusive sensing from fixed vantage points, thus constituting a unique tool for underwater environmental monitoring.

Enhancing remote target classification in hyperspectral imaging using graph attention neural network

Enhancing remote target classification in hyperspectral imaging using graph attention neural network

A feature selection method for multimodal multispectral LiDAR sensing

Optical remote sensing techniques can indicate the properties of objects by observing different modalities (physical quantities) of the backscattered light at different optical wavelengths. Established examples are reflectance, fluorescence, Raman, or depolarization spectroscopy. LiDAR sensing, on the other hand, allows acquiring the geometry of objects by measuring the propagation delay of optical probing signals. Multimodal multispectral (MM) LiDAR combines these capabilities and extends conventional monochromatic LiDAR in both spectral and modal dimensions within a single instrument, thus enriching point cloud data with non-geometric information. The potentially high dimension of MM LiDAR data, however, poses significant challenges for instrumental design, data acquisition, and data processing. MM LiDAR data are structured as several or all modalities are available in each of the spectral channels. The above challenges can thus be mitigated by feature selection (FS), if the structure of the features is taken into account, i.e., if entire spectral channels or entire modalities are selected or omitted. Herein, we focus on the feature selection method for MM LiDAR and propose a multiclass group feature selection algorithm (MGSVM FS) consisting of a structural sparsity-based embedded feature selection method with an all-in-one support vector machine (SVM). It tackles jointly the challenges arising from the high dimension of the MM data and the need for a multiclass classification task while exploiting the structure of the MM data. In addition, we introduce a complete workflow for evaluating the feature selection and for decision-making. We apply the framework to selecting an optimum spectral and modal configuration for remote material classification using an experimental MM LiDAR system that provides reflectance, distance, and degree of linear polarization in 28 spectral channels of 10 nm width. For the experimental investigation, we use MM LiDAR data obtained in a controlled lab environment from thirty specimens of four material classes relevant for construction. Using all three modalities, we find a configuration with only 3 spectral channels that achieves a classification mean-F1 score of 100% within this small dataset. Similar classification performance can also be achieved with only two modalities when using more spectral channels. MGSVM FS improves the classification mean-F1 score by up to 25% as compared to random selection and outperforms two other commonly used filter and embedded feature selection methods, in this application example. The proposed group feature selection algorithm and decision-making are useful for MM LiDAR, providing a link between instrumental design, data acquisition, and data processing. However, they are also transferable to other application fields related to multiclass classification, regression, and knowledge discovery, with features structured in groups. The collected MM feature dataset, the MGSVM FS algorithm, and the evaluation pipeline are accessible online.11https://github.com/yuhan-yhyh/Dataset_Code_MGSVM-FS-MM-LiDAR.git.

Remote Disease Diagnosis through IoMT-Enhanced Blood Cell Classification with Deep Learning

Background For the purpose of diagnosing diseases and developing treatment plans, blood cell pictures must be accurately classified. This procedure can be greatly enhanced by automated systems that make use of deep learning and the Internet of Medical Things (IoMT). Objective In order to improve illness detection and increase healthcare accessibility, this work suggests an IoMT-based system for remote blood cell picture transmission and classification utilizing deep learning algorithms. Methods High-resolution pictures of blood cells are captured by an IoMT tiny camera and wirelessly sent to a cloud-based infrastructure. The blood cells are divided into groups according to a, deeplearning classification algorithm, including neutrophils, lymphocytes, monocytes, and eosinophils. Results The IoMT-enabled system excels in transmitting and analyzing blood cell images, achieving precise classification. Utilizing deep learning models with multi-scale feature extraction and attention mechanisms, the system demonstrates robust performance. Numerical results showcase a high accuracy of approximately 97.21%, along with noteworthy precision, recall, and F1 scores for individual blood cell classes. Eosinophil, Lymphocyte, Monocyte, and Neutrophil classes exhibit strong performance metrics, emphasizing the system's effectiveness in accurate blood cell classification. Conclusion By combining IoMT and deep learning with blood cell image analysis, diagnostic accessibility and efficiency are improved. The suggested approach has the potential to completely transform healthcare by facilitating prompt interventions, individualized treatment regimens, and better patient outcomes. It is essential to continuously enhance and validate the system in order to maximize its efficacy and dependability in a variety of healthcare settings.

The reliability of remote diagnostic hearing assessment in school-entry-aged children

Objective To explore the feasibility of remote diagnostic hearing testing within schools, for school-entry aged children (4–7 years). Design A prospective cross-sectional comparative study design was used to establish the reliability of remote hearing assessment by comparing traditional face-to-face pure tone audiometry (PTA) with remote PTA, traditional otoscopy with remote video otoscopy and interpretation of tympanograms in-person with remote analysis. Study Sample 464 school-entry year level children from 18 schools. Results Air conduction thresholds established by remote testing were within a clinically acceptable range of ±10 dB HL from traditional thresholds in 98% of cases at 1 kHz and 97.8% at 4 kHz. Bone conduction thresholds varied by ≤10 dB HL in 100% of cases at 1 kHz and 95.7% at 4 kHz. Remote otoscopy and tympanometry interpretation had 78.6% agreement (moderate; κ = 0.6) and 92.2% agreement (almost perfect; κ = 0.9), respectively. Conclusions We found there to be good reliability between face-to-face PTA and remote PTA performed within the school environment. Furthermore, assessment of the outer and middle ear via video otoscopy and remote tympanometry classification were both viable. Remote diagnostic hearing tests in schools can extend reach of paediatric services in rural areas.

Multi-Sensor Classification Framework of Urban Vegetation for Improving Ecological Services Management

Recent climatic changes have profoundly impacted the urban microclimate, exposing city dwellers to harsh living conditions. One effective approach to mitigating these events involves incorporating more green infrastructure into the cityscape. The ecological services provided by urban vegetation play a crucial role in enhancing the sustainability and livability of cities. However, monitoring urban vegetation and accurately estimating its status pose challenges due to the heterogeneous nature of the urban environment. In response to this, the current study proposes utilizing a remote sensing-based classification framework to enhance data availability, thereby improving practices related to urban vegetation management. The aim of the current research is to explore the spatial pattern of vegetation and enhance the classification of tree species within diverse and complex urban environments. This study combines various remote sensing observations to enhance classification capabilities. High-resolution colored rectified aerial photographs, LiDAR-derived products, and hyperspectral data are merged and analyzed using advanced classifier methods, specifically partial least squares-discriminant analysis (PLS-DA) and object-based image analysis (OBIA). The OBIA method demonstrates an impressive overall accuracy of 95.30%, while the PLS-DA model excels with a remarkable overall accuracy of 100%. The findings validate the efficacy of incorporating OBIA, aerial photographs, LiDAR, and hyperspectral data in improving tree species classification and mapping within the context of PLS-DA. This classification framework holds significant potential for enhancing management practices and tools, thereby optimizing the ecological services provided by urban vegetation and fostering the development of sustainable cities.

Determination of asbestos cement rooftop surface composition using regression analysis and hyper-spectral reflectance data in the visible and near-infrared ranges

The effects of asbestos on human health have spurred numerous studies examining its risks in urban environments. Recent works have shifted towards less-invasive techniques for remote detection and classification of asbestos-cement. In this context, this study combines visible (VIS) and near-infrared (NIR) reflectance data collected in-situ with reference signals from the USGS spectral library, utilizing optimized regression analysis to determine the surface composition of corrugated asbestos-cement rooftops. An outlier filter was successfully implemented to enhance the accuracy of regression calculations, achieving a high level of agreement with actual field observations. The regression analysis revealed varying proportions of weathered cement, hazardous asbestos fibers (specifically chrysotile and cummingtonite), and biological growth (such as lichens and moss). These results are consistent with previous research on the composition of asbestos-cement rooftops, including a comparable field study and XRD analysis conducted in 2019. This underscores the importance of using regression analysis, preceded by an outlier filtering step, on VIS and NIR reflectance data to ascertain the surface composition of asbestos-cement rooftops. This methodology holds potential for application to larger hyperspectral datasets across more extensive sample surfaces and areas.

Integrating NDVI-Based Within-Wetland Vegetation Classification in a Land Surface Model Improves Methane Emission Estimations

Earth system models (ESMs) are a common tool for estimating local and global greenhouse gas emissions under current and projected future conditions. Efforts are underway to expand the representation of wetlands in the Energy Exascale Earth System Model (E3SM) Land Model (ELM) by resolving the simultaneous contributions to greenhouse gas fluxes from multiple, different, sub-grid-scale patch-types, representing different eco-hydrological patches within a wetland. However, for this effort to be effective, it should be coupled with the detection and mapping of within-wetland eco-hydrological patches in real-world wetlands, providing models with corresponding information about vegetation cover. In this short communication, we describe the application of a recently developed NDVI-based method for within-wetland vegetation classification on a coastal wetland in Louisiana and the use of the resulting yearly vegetation cover as input for ELM simulations. Processed Harmonized Landsat and Sentinel-2 (HLS) datasets were used to drive the sub-grid composition of simulated wetland vegetation each year, thus tracking the spatial heterogeneity of wetlands at sufficient spatial and temporal resolutions and providing necessary input for improving the estimation of methane emissions from wetlands. Our results show that including NDVI-based classification in an ELM reduced the uncertainty in predicted methane flux by decreasing the model’s RMSE when compared to Eddy Covariance measurements, while a minimal bias was introduced due to the resampling technique involved in processing HLS data. Our study shows promising results in integrating the remote sensing-based classification of within-wetland vegetation cover into earth system models, while improving their performances toward more accurate predictions of important greenhouse gas emissions.

Study of methods for searching and localizing objects in images from aircraft using convolutional neural networks

The use of unmanned and manned aerial vehicles for remote object localization and classification is very common. These methods are used in various systems, ranging from territory surveys to law enforcement. Methods of object localization and classification using neural networks require a detailed study and research of the quality of their work on data that has certain specifics, such as vehicle detection. The use of neural networks to detect certain types of objects using images obtained from aircraft can also help in the study of hard-to-reach locations. Therefore, the main subject of this paper is the localization and classification of objects in images obtained using digital cameras mounted on aircraft. The main focus is on determining the accuracy of object localization and detection using selected types of neural networks, which are the most important indicators of neural network efficiency. The speed of a neural network is also an equally important characteristic as it directly affects its ability to be used in tasks that require fast object localization, such as video surveillance or automated car control systems. The main goal of this study is to study the accuracy of object localization and classification in images obtained with the help of cameras mounted on aircraft, as well as to study the speed of neural networks and determine the effectiveness of their application in real-world conditions. The objectives of this study are to train YOLO v5, SSD, and Faster RCNNs on the VisDrone dataset and to further study them on the vehicle localization dataset. The main goal of this work is to obtain statistics on the performance of neural networks trained on the VisDrone dataset. On the basis of the obtained statistics, conclusions are drawn about the effectiveness of the considered neural networks. The conclusions are drawn by considering the speed of the model, localization (IoU), and classification (Precision, Recall) metrics. Possible directions for further development of the topic under study are presented as conclusions.

Supervised Classification Accuracy Assessment Using Remote Sensing and Geographic Information System

Assessing the accuracy of classification algorithms is paramount as it provides insights into reliability and effectiveness in solving real-world problems. Accuracy examination is essential in any remote sensing-based classification practice, given that classification maps consistently include misclassified pixels and classification misconceptions. In this study, two imaginary satellites for Duhok province, Iraq, were captured at regular intervals, and the photos were analyzed using spatial analysis tools to provide supervised classifications. Some processes were conducted to enhance the categorization, like smoothing. The classification results indicate that Duhok province is divided into four classes: vegetation cover, buildings, water bodies, and bare lands. During 2013-2022, vegetation cover increased from 63% in 2013 to 66% in 2022; buildings roughly increased by 1% to 3% yearly; water bodies showed a decrease of 2% to 1%; the amount of unoccupied land showed a decrease from 34% to 30%. Therefore, the classification accuracy was assessed using the approach of comparison with field data; the classification accuracy was about 85%.

Accuracy Measurement of Hyperspectral Image Classification in Remote Sensing with the Light Spectrum-based Affinity Propagation Clustering-based Segmentation

The area of remote sensing and computer vision includes the challenge of hyperspectral image classification. It entails grouping pixels in hyperspectral pictures into several classes according to their spectral signature. Hyperspectral photographs are helpful for a variety of applications, including vegetation study, mineral mapping, and mapping urban land use, since they include information on an object's reflectance in hundreds of small, contiguous wavelength bands. This task's objective is to correctly identify and categorize several item categories in the image. Many approaches have been stated by several researchers in this field to enhance the accuracy of the segmentation and accuracy. However, fails to attain the optimal accuracy due to the intricate nature of the images. To tackle these issues, we propose a novel Modified Extreme Learning machine (M-ELM) approach for the credible hyperspectral image classification outcomes with the publicly available Kaggle datasets. Before the classification, the input images are segmented using the Light Spectrum-based modified affinity propagation clustering technique (LSO-MAPC). In the beginning, the images are pre-processed using the non-linear diffusion partial differential equations technique which effectively pre-processed the image spatially. Experiments are effectuated to analyze the performance of the proposed method and compared it with state-of-art works in a quantitative way. The proposed approach ensures a classification accuracy of 96%.

FPGA-based remote target classification in hyperspectral imaging using multi-graph neural network

Hyperspectral imagery (HSI) is widely used in remote sensing for target classification; however, its accurate classification remains challenging due to the scarcity of labeled data. Graph Neural Networks (GNNs) have emerged as a popular method for semi-supervised classification, attracting significant interest in the context of HSI analysis. Nevertheless, conventional GNN-based approaches often rely on a single graph filter to extract HSI characteristics, failing to fully exploit the potential benefits of different graph filters. Additionally, oversmoothing issues plague classical GNNs, further affecting classification performance. To address these drawbacks, we propose a novel approach called Spectral and Autoregressive Moving Average Graph Filter for the Multi-Graph Neural Network (SAM-GNN). This approach leverages two distinct graph filters: one specialized in extracting the spectral characteristics of nodes and the other effectively suppressing graph distortion. Through extensive evaluations, we compare the performance of SAM-GNN with other state-of-the-art methods, employing metrics such as overall accuracy (OA), individual class accuracy (IA), and Kappa coefficient (KC). The results shows that the SAM-GNN provides an improvement in KC, IA, and OA of 6.71%, 5.7%, and 3.93% for the Pavia University dataset and 4.67%, 3.67%, and 3.49% for the Cuprite dataset respectively. Furthermore, we implement SAM-GNN on the Virtex-7 field-programmable gate array (FPGA), demonstrating that the method achieves highly accurate target localization results, bringing us closer to real-world applications in HSI classification.

Internet of things-based remote monitoring and classification of <i>Spinacia oleracea</i> leaf disease using deep learning approach

Internet of things-based remote monitoring and classification of <i>Spinacia oleracea</i> leaf disease using deep learning approach

Internet of Things-Based Remote Monitoring and Classification of Spinacia Oleracea Leaf Disease Using Deep Learning Approach

Internet of Things-Based Remote Monitoring and Classification of Spinacia Oleracea Leaf Disease Using Deep Learning Approach

Model-informed classification of broadband acoustic backscatter from zooplankton in an in situ mesocosm

Abstract Classification of zooplankton to species with broadband echosounder data could increase the taxonomic resolution of acoustic surveys and reduce the dependence on net and trawl samples for ‘ground truthing’. Supervised classification with broadband echosounder data is limited by the acquisition of validated data required to train machine learning algorithms (‘classifiers’). We tested the hypothesis that acoustic scattering models could be used to train classifiers for remote classification of zooplankton. Three classifiers were trained with data from scattering models of four Arctic zooplankton groups (copepods, euphausiids, chaetognaths, and hydrozoans). We evaluated classifier predictions against observations of a mixed zooplankton community in a submerged purpose-built mesocosm (12 m3) insonified with broadband transmissions (185–255 kHz). The mesocosm was deployed from a wharf in Ny-Ålesund, Svalbard, during the Arctic polar night in January 2022. We detected 7722 tracked single targets, which were used to evaluate the classifier predictions of measured zooplankton targets. The classifiers could differentiate copepods from the other groups reasonably well, but they could not differentiate euphausiids, chaetognaths, and hydrozoans reliably due to the similarities in their modelled target spectra. We recommend that model-informed classification of zooplankton from broadband acoustic signals be used with caution until a better understanding of in situ target spectra variability is gained.

Remote sensing-based scene classification by feature fusion and extraction with ensemble classifier employing machine learning approaches

This research focuses on scene segmentation in remotely sensed images within the field of Remote Sensing Image Scene Understanding (RSISU). Leveraging recent advancements in Deep Learning (DL), particularly Residual Neural Networks (RESNET-50 and RESNET-101), and the research proposes a methodology involving feature fusing, extraction, and classification for categorizing remote sensing images. The approach employs a dataset from the University of California Irvine (UCI) comprising twenty-one groups of pictures. The images undergo pre-processing, feature extraction using the mentioned DL frameworks, and subsequent categorization through an ensemble classification structure combining Kernel Extreme Learning Machine (KELM) and Support Vector Machine (SVM). The paper concludes with optimal results achieved through performance and comparison analyses.

Enhanced Ship/Iceberg Classification in SAR Images Using Feature Extraction and the Fusion of Machine Learning Algorithms

Drifting icebergs present significant navigational and operational risks in remote offshore regions, particularly along the East Coast of Canada. In such areas with harsh weather conditions, traditional methods of monitoring and assessing iceberg-related hazards, such as aerial reconnaissance and shore-based support, are often unfeasible. As a result, satellite-based monitoring using Synthetic Aperture Radar (SAR) imagery emerges as a practical solution for timely and remote iceberg classifications. We utilize the C-CORE/Statoil dataset, a labeled dataset containing both ship and iceberg instances. This dataset is derived from dual-polarized Sentinel-1. Our methodology combines state-of-the-art deep learning techniques with comprehensive feature selection. These features are coupled with machine learning algorithms (neural network, LightGBM, and CatBoost) to achieve accurate and efficient classification results. By utilizing quantitative features, we capture subtle patterns that enhance the model’s discriminative capabilities. Through extensive experiments on the provided dataset, our approach achieves a remarkable accuracy of 95.4% and a log loss of 0.11 in distinguishing icebergs from ships in SAR images. The introduction of additional ship images from another dataset can further enhance both accuracy and log loss results to 96.1% and 0.09, respectively.