AIS Data Driven CNN-BiGRU Model for Ship Target Classification

The lung is a muscular organ which is the size of a clenched human fist and is responsible for blood circulation. To create a bottleneck classification model we required the minimization of the classification of the entire raw data. Although lung disease is the general name given to diseases affecting the lung, there are, in fact, an enormous range of diseases, including chronic obstructive pulmonary disease (COPD), lung pneumonia, lung cancer and so on depending upon the circulation of blood and the shortness of breath throughout the body. Hence, this research work initiate a hybrid method named HCFFSCACO (Hierarchical clustering Fuzzy features subset classifier with ant colony optimization) for identifying appropriate feature subsets related to the target class and given to the classifier model to enhance the performance of health monitoring and management (HMM). This is made feasible by incorporating a feature selection method with a classification model using the UCI lung dataset. Qualitative assessment of the proposed hierarchical clustering feature subsets classification models and privacy preserving mechanism has been made with classification accuracy and running time, respectively. The presented systems provide compromised results over existing methods, according on statistical analysis of accuracy values and computational time.

Real-world datasets are heavily skewed where some classes are significantly outnumbered by the other classes. In these situations, machine learning algorithms fail to achieve substantial efficacy while predicting these underrepresented instances. To solve this problem, many variations of synthetic minority oversampling methods (SMOTE) have been proposed to balance datasets which deal with continuous features. However, for datasets with both nominal and continuous features, SMOTE-NC is the only SMOTE-based oversampling technique to balance the data. In this paper, we present a novel minority oversampling method, SMOTE-ENC (SMOTE—Encoded Nominal and Continuous), in which nominal features are encoded as numeric values and the difference between two such numeric values reflects the amount of change of association with the minority class. Our experiments show that classification models using the SMOTE-ENC method offer better prediction than models using SMOTE-NC when the dataset has a substantial number of nominal features and also when there is some association between the categorical features and the target class. Additionally, our proposed method addressed one of the major limitations of the SMOTE-NC algorithm. SMOTE-NC can be applied only on mixed datasets that have features consisting of both continuous and nominal features and cannot function if all the features of the dataset are nominal. Our novel method has been generalized to be applied to both mixed datasets and nominal-only datasets.

Support vector machine (SVM) is a machine learning algorithm for learning classification and regression models. To build intuition, we will consider the case of learning a classification model with SVM. Given a dataset with two target classes that are linearly separable, it turns out that there exists an infinite number of lines that can discriminate between the two classes (see Figure 22-1). The goal of the SVM is to find the best line that separates the two classes. In higher dimensions, this line is called a hyperplane.

Target encoding is an effective technique to boost performance of classical and deep neural networks based classification models. However, the existing target encoding approaches require significant increase in the learning capacity, thus demand higher computation power and more training data. In this paper, we present a novel and efficient target encoding method, Inter-class Ambiguity Driven Multi-hot Target Encoding (MUTE), to improve both generalizability and robustness of a classification model by understanding the inter-class characteristics of a target dataset. By evaluating ambiguity between the target classes in a dataset, MUTE strategically optimizes the Hamming distances among target encoding. Such optimized target encoding offers higher classification strength for neural network models with negligible computation overhead and without increasing the model size. When MUTE is applied to the popular image classification networks and datasets, our experimental results show that MUTE offers better generalization and defense against the noises and adversarial attacks over the existing solutions.

Intelligent video surveillance system has good application prospect in security field, the detection, tracking and classification of target are key parts of the system. To achieve target detection, this paper applies Bayesian model for tracking and classification of target, Bayesian theorem provides probability judgment means, the hypothesis examination variables follow probability distribution, these probability and observation data are inferenced in order to make the best decisions, and which can track moving targets which are in the prospects region and the background. The results have shown that the system can be used for indoor tracking of target detection and classification, and can quickly detect, track and make classification of moving targets.

There is high risk of collision between passenger ferries and other nearby ships in Inland River when the ferries are crossing the channel. Some serious accidents might happen and result in catastrophic fatalities. Therefore, researches on the behavior feature of ferries are performed in this paper. Most of the information related to ship's encountering can be extracted from AIS data, which provides with effective assistance to the investigation on the behavior features of ferries. As a result, the AIS data of ships in Jiangsu section of the Yangtze River is introduced, and the linear interpolation is applied after data preprocessing to obtain the navigation data of other ships near ferries at the same time. Based on this, some statistical analysis, including the encounter situation, ship type, size, navigation speed and course when the passenger ferries are encountered with target ships, are analyzed so that the detailed navigation features about the crossing behaviors can be identified in more detail. This study presents a theoretical basis for ensuring the navigation safety of passenger ferries and reducing the probability of accidents.

High-resolution space-borne synthetic aperture radar (SAR) images are important sources of information for marine monitoring and target recognition. For inshore port scenes, there is a wide variety of ship targets; moreover, as the scattering intensity of the land background is much higher than that of the sea surface, similar structures on land in SAR images can cause significant interference to the identification of ship targets, resulting in a high false alarm rate and false recognition rate. Therefore, the classification of ship targets in SAR images still requires professional interpretation and correction to eliminate possible false alarms and recognitions. In this paper, by imitating the basic structural requirements for manual discrimination of target categories, a space-borne SAR high-resolution inshore ship classification and correction method of “Pre-recognition – Fine Recognition – Matching Correction” is designed based on the YOLO recognition network. Taking three types of typical ship targets as examples, a high-resolution SAR inshore ship target dataset and a high-resolution SAR ship component architecture dataset are constructed to train the pre-recognition network and the fine recognition network respectively. The latter is based on the former to recognize the component composition of possible targets, and forms a component arrangement architecture vector. The final recognition results are obtained by matching and correcting the similarity distance with the three types of targets. This method can reduce the false alarm rate in ship target recognition in complex inshore port scenes to less than 10% and correct some false recognition results, thereby improving the accuracy of ship target identification and reducing the cost for manual correction.

Deep learning models are being applied to more and more use cases with astonishing success stories, but how do they perform in the real world? Models are typically tested on specific cleaned data sets, but when deployed in the real world, the model will encounter unexpected, out-of-distribution (OOD)data. To investigate the impact of OOD radiographs on existing chest x-ray classification models and to increase their robustness against OODdata. The study employed the commonly used chest x-ray classification model, CheXnet, trained on the chest x-ray 14 data set, and tested its robustness against OOD data using three public radiography data sets: IRMA, Bone Age, and MURA, and the ImageNet data set. To detect OOD data for multi-label classification, we proposed in-distribution voting (IDV). The OOD detection performance is measured across data sets using the area under the receiver operating characteristic curve (AUC) analysis and compared with Mahalanobis-based OOD detection, MaxLogit, MaxEnergy, self-supervised OOD detection (SS OOD), andCutMix. Without additional OOD detection, the chest x-ray classifier failed to discard any OOD images, with an AUC of 0.5. The proposed IDV approach trained on ID (chest x-ray 14) and OOD data (IRMA and ImageNet) achieved, on average, 0.999 OOD AUC across the three data sets, surpassing all other OOD detection methods. Mahalanobis-based OOD detection achieved an average OOD detection AUC of 0.982. IDV trained solely with a few thousand ImageNet images had an AUC 0.913, which was considerably higher than MaxLogit (0.726), MaxEnergy (0.724), SS OOD (0.476), and CutMix (0.376). The performance of all tested OOD detection methods did not translate well to radiography data sets, except Mahalanobis-based OOD detection and the proposed IDV method. Consequently, training solely on ID data led to incorrect classification of OOD images as ID, resulting in increased false positive rates. IDV substantially improved the model's ID classification performance, even when trained with data that will not occur in the intended use case or test set (ImageNet), without additional inference overhead or performance decrease in the target classification. The corresponding code is available at https://gitlab.lrz.de/IP/a-knee-cannot-have-lung-disease.

BackgroundPresence of multimodal brain graphs derived from different neuroimaging modalities is inarguably one of the most critical challenges in building unified classification models that can be trained and tested on any brain graph regardless of its size and the modality it was derived from. Existing methodsOne solution is to learn a model for each modality independently, which is cumbersome and becomes more time-consuming as the number of modalities increases. Another traditional solution is to build a model inputting multimodal brain graphs for the target prediction task; however, this is only applicable to datasets where all samples have joint neuro-modalities. New methodIn this paper, we propose to build a unified brain graph classification model trained on unpaired multimodal brain graphs, which can classify any brain graph of any size. This is enabled by incorporating a graph alignment step where all multi-modal graphs of different sizes and heterogeneous distributions are mapped to a common template graph. Next, we design a graph alignment strategy to the target fixed-size template and further apply linear discriminant analysis (LDA) to the aligned graphs as a supervised dimensionality reduction technique for the target classification task. ResultsWe tested our method on unpaired autistic and healthy brain connectomes derived from functional and morphological MRI datasets (two modalities). ConclusionOur results showed that our unified model method not only has great promise in solving such a challenging problem but achieves comparable performance to models trained on each modality independently.

The rapid serial visual presentation-based brain–computer interface (RSVP-BCI) system achieves the recognition of target images by extracting event-related potential (ERP) features from electroencephalogram (EEG) signals and then building target classification models. Currently, how to reduce the training and calibration time for classification models across different subjects is a crucial issue in the practical application of RSVP. To address this issue, a zero-calibration (ZC) method termed Attention-ProNet, which involves meta-learning with a prototype network integrating multiple attention mechanisms, was proposed in this study. In particular, multiscale attention mechanisms were used for efficient EEG feature extraction. Furthermore, a hybrid attention mechanism was introduced to enhance model generalization, and attempts were made to incorporate suitable data augmentation and channel selection methods to develop an innovative and high-performance ZC RSVP-BCI decoding model algorithm. The experimental results demonstrated that our method achieved a balance accuracy (BA) of 86.33% in the decoding task for new subjects. Moreover, appropriate channel selection and data augmentation methods further enhanced the performance of the network by affording an additional 2.3% increase in BA. The model generated by the meta-learning prototype network Attention-ProNet, which incorporates multiple attention mechanisms, allows for the efficient and accurate decoding of new subjects without the need for recalibration or retraining.

Systems that rely on Face Detection have gained great importance ever, since large-scale databases of thousands of face images are collected from several sources. Thus, the use of an outperforming face detector becomes a challenging problem. Different classification models have been studied and applied for face detection. However, such models involve large scale datasets, which requires huge memory and enormous amount of training time. Therefore, in this paper, we investigate the potency of incrementally projecting data in low variance directions. In fact, in one-class classification, the low variance directions in the training data carry crucial information to build a good model of the target class. On the other hand, incremental learning is known to be powerful, when dealing with dynamic data. We performed extensive tests on human faces, and comparative experiments have been carried out to show the effectiveness and superiority of our proposed method over other face detection methods.

The explosive rise of malware poses risks to Android developers and organization regarding security lapses and monetary losses. The dynamic nature, changing complexity and behavior over time, and increasing velocity and volume make it challenging for the malware protection community to provide a robust and reliable protection system. Due to these characteristics, conventional Android malware detection techniques, such as signature-based and battery-monitoring, cannot detect futuristic malware. Current research exploiting deep learning methods shows excellent performance compared to conventional and machine learning methods. However, the majority of the techniques are proposed for only binary classification. These classification models are tested on customized datasets. They do not provide the model’s effectiveness in terms of generalization, as the model’s accuracy might be good for some malware classes. Hence, providing a practical, robust, stable, and reliable malware model is still an open issue. Therefore, in this work, we propose an Attention-based deep learning model to detect categorical malware classes. The attention-based deep learning mechanism learns the malicious behavior of target classes. The attention mechanism filters and extracts the relevant information more effectively by focusing on the specific keywords in a target sample.

The classification of ship targets in remote sensing images holds significant application value in fields such as marine monitoring and national defence. Although existing research has yielded considerable achievements in ship classification, current methods struggle to distinguish highly similar ship categories for fine-grained classification tasks due to a lack of targeted design. Specifically, they exhibit the following shortcomings: limited ability to extract locally discriminative features; inadequate fusion of features at high and low levels of representation granularity; and sensitivity of model performance to background noise. To address this issue, this paper proposes a fine-grained classification framework for ship targets in remote sensing images based on Multi-Granularity Attention Network (MA-Net), specifically designed to tackle the aforementioned three major challenges encountered in fine-grained classification tasks for ship targets in remote sensing. This framework first performs multi-level feature extraction through a backbone network, subsequently introducing an Adaptive Local Feature Attention (ALFA) module. This module employs dynamic overlapping region segmentation techniques to assist the network in learning spatial structural combinations, thereby optimising the representation of local features. Secondly, a Dynamic Multi-Granularity Feature Fusion (DMGFF) module is designed to dynamically fuse feature maps of varying representational granularities and select key attribute features. Finally, a Feature-Based Data Augmentation (FBDA) method is developed to effectively highlight target detail features, thereby enhancing feature expression capabilities. On the public FGSC-23 and FGSCR-42 datasets, MA-Net attains top-performing accuracies of 93.12% and 98.40%, surpassing the previous best methods and establishing a new state of the art for fine-grained classification of ship targets in remote sensing images.

On the usefulness of one-class classifier ensembles for decomposition of multi-class problems

As an important part of modern marine monitoring systems, ship target identification has important significance in maintaining marine rights and monitoring maritime traffic. With the development of artificial intelligence technology, image detection and recognition based on deep learning methods have become the most popular and practical method. In this paper, two deep learning algorithms, the Mask R-CNN algorithm and the Faster R-CNN algorithm, are used to build ship target feature extraction and recognition models based on deep convolutional neural networks. The established models were compared and analyzed to verify the feasibility of target detection algorithms. In this study, 5748 remote sensing maps were selected as the dataset for experiments, and two algorithms were used to classify and extract warships and civilian ships. Experiments showed that for the accuracy of ship identification, Mask R-CNN and Faster R-CNN reached 95.21% and 92.76%, respectively. These results demonstrated that the Mask R-CNN algorithm achieves pixel-level segmentation. Compared with the Faster R-CNN algorithm, the obtained target detection effect is more accurate, and the performance in target detection and classification is better, which reflects the great advantage of pixel-level recognition.