Search Results (27)

Travel mode identification (TMI) plays a crucial role in intelligent transportation systems by accurately identifying travel modes from Global Positioning System (GPS) trajectory data. Given that trajectory data inherently exhibit spatial and kinematic patterns that complement each other, recent TMI methods generally combine these characteristics through image-based projections or direct concatenation. However, such approaches achieve only shallow fusion of these two types of features and cannot effectively align them into a shared latent space. To overcome this limitation, we introduce multi-view contrastive fusion (MVCF)-TMI, a novel TMI framework that enhances identification accuracy and model generalizability by aligning spatial and kinematic views through multi-view contrastive learning. Our framework employs multi-view learning to separately extract spatial and kinematic features, followed by an inter-view contrastive loss to optimize feature alignment in a shared subspace. This approach enables cross-view semantic understanding and better captures complementary information across different trajectory representations. Extensive experiments show that MVCF-TMI outperforms baseline methods, achieving 86.45% accuracy on the GeoLife dataset. The model also demonstrates strong generalization by transferring knowledge from pretraining on the large-scale GeoLife dataset to the smaller SHL dataset. Full article

Skin cancer is among the most prevalent cancers globally, emphasizing the need for early detection and accurate diagnosis to improve outcomes. Traditional diagnostic methods, based on visual examination, are subjective, time-intensive, and require specialized expertise. Current artificial intelligence (AI) approaches for skin cancer detection face challenges such as computational inefficiency, lack of interpretability, and reliance on standalone CNN architectures. To address these limitations, this study proposes a comprehensive pipeline combining transfer learning, feature selection, and machine-learning algorithms to improve detection accuracy. Multiple pretrained CNN models were evaluated, with Xception emerging as the optimal choice for its balance of computational efficiency and performance. An ablation study further validated the effectiveness of freezing task-specific layers within the Xception architecture. Feature dimensionality was optimized using Particle Swarm Optimization, reducing dimensions from 1024 to 508, significantly enhancing computational efficiency. Machine-learning classifiers, including Subspace KNN and Medium Gaussian SVM, further improved classification accuracy. Evaluated on the ISIC 2018 and HAM10000 datasets, the proposed pipeline achieved impressive accuracies of 98.5% and 86.1%, respectively. Moreover, Explainable-AI (XAI) techniques, such as Grad-CAM, LIME, and Occlusion Sensitivity, enhanced interpretability. This approach provides a robust, efficient, and interpretable solution for automated skin cancer diagnosis in clinical applications. Full article

Machine learning (ML) methods are increasingly being applied to analyze biological signals. For example, ML methods have been successfully applied to the human electroencephalogram (EEG) to classify neural signals as pathological or non-pathological and to predict working memory performance in healthy and psychiatric patients. ML approaches can quickly process large volumes of data to reveal patterns that may be missed by humans. This study investigated the accuracy of ML methods at classifying the brain’s electrical activity to cognitive events, i.e., event-related brain potentials (ERPs). ERPs are extracted from the ongoing EEG and represent electrical potentials in response to specific events. ERPs were evoked during a visual Go/NoGo task. The Go/NoGo task requires a button press on Go trials and response withholding on NoGo trials. NoGo trials elicit neural activity associated with inhibitory control processes. We compared the accuracy of six ML algorithms at classifying the ERPs associated with each trial type. The raw electrical signals were fed to all ML algorithms to build predictive models. The same raw data were then truncated in length and fitted to multiple dynamic state space models of order $n_x$ using a continuous-time subspace-based system identification algorithm. The $4n_x$ numerator and denominator parameters of the transfer function of the state space model were then used as substitutes for the data. Dimensionality reduction simplifies classification, reduces noise, and may ultimately improve the predictive power of ML models. Our findings revealed that all ML methods correctly classified the electrical signal associated with each trial type with a high degree of accuracy, and accuracy remained high after parameterization was applied. We discuss the models and the usefulness of the parameterization. Full article

Acute lymphoblastic leukemia (ALL) is a lethal blood cancer that is characterized by an abnormal increased number of immature lymphocytes in the blood or bone marrow. For effective treatment of ALL, early assessment of the disease is essential. Manual examination of stained blood smear images is current practice for initially screening ALL. This practice is time-consuming and error-prone. In order to effectively diagnose ALL, numerous deep-learning-based computer vision systems have been developed for detecting ALL in blood peripheral images (BPIs). Such systems extract a huge number of image features and use them to perform the classification task. The extracted features may contain irrelevant or redundant features that could reduce classification accuracy and increase the running time of the classifier. Feature selection is considered an effective tool to mitigate the curse of the dimensionality problem and alleviate its corresponding shortcomings. One of the most effective dimensionality-reduction tools is principal component analysis (PCA), which maps input features into an orthogonal space and extracts the features that convey the highest variability from the data. Other feature selection approaches utilize evolutionary computation (EC) to search the feature space and localize optimal features. To profit from both feature selection approaches in improving the classification performance of ALL, in this study, a new hybrid deep-learning-based feature engineering approach is proposed. The introduced approach integrates the powerful capability of PCA and particle swarm optimization (PSO) approaches in selecting informative features from BPI mages with the power of pre-trained CNNs of feature extraction. Image features are first extracted through the feature-transfer capability of the GoogleNet convolutional neural network (CNN). PCA is utilized to generate a feature set of the principal components that covers 95% of the variability in the data. In parallel, bio-inspired particle swarm optimization is used to search for the optimal image features. The PCA and PSO-derived feature sets are then integrated to develop a hybrid set of features that are then used to train a Bayesian-based optimized support vector machine (SVM) and subspace discriminant ensemble-learning (SDEL) classifiers. The obtained results show improved classification performance for the ML classifiers trained by the proposed hybrid feature set over the original PCA, PSO, and all extracted feature sets for ALL multi-class classification. The Bayesian-optimized SVM trained with the proposed hybrid PCA-PSO feature set achieves the highest classification accuracy of 97.4%. The classification performance of the proposed feature engineering approach competes with the state of the art. Full article

Batch processes in the biopharmaceutical and chemical manufacturing industries often develop new products to meet changing market demands. When the dynamic models of these new products are trained, dynamic modeling with limited data for each product can lead to inaccurate results. One solution is to extract useful knowledge from past historical production data that can be applied to the product of a new grade. In this way, the model can be built quickly without having to wait for additional modeling data. In this study, a subspace identification combined common feature learning scheme is proposed to quickly learn a model of a new grade. The proposed modified state-space model contains common and special parameter matrices. Past batch data can be used to train common parameter matrices. Then, the parameters can be directly transferred into a new SID model for a new grade of the product. The new SID model can be quickly well trained even though there is a limited batch of data. The effectiveness of the proposed algorithm is demonstrated in a numerical example and a case of an industrial penicillin process. In these cases, the proposed common feature extraction for the SID learning framework can achieve higher performance in the multi-input and multi-output batch process regression problem. Full article

Skin cancer classification is a complex and time-consuming task. Existing approaches use segmentation to improve accuracy and efficiency, but due to different sizes and shapes of lesions, segmentation is not a suitable approach. In this research study, we proposed an improved automated system based on hybrid and optimal feature selections. Firstly, we balanced our dataset by applying three different transformation techniques, which include brightness, sharpening, and contrast enhancement. Secondly, we retrained two CNNs, Darknet53 and Inception V3, using transfer learning. Thirdly, the retrained models were used to extract deep features from the dataset. Lastly, optimal features were selected using moth flame optimization (MFO) to overcome the curse of dimensionality. This helped us in improving accuracy and efficiency of our model. We achieved 95.9%, 95.0%, and 95.8% on cubic SVM, quadratic SVM, and ensemble subspace discriminants, respectively. We compared our technique with state-of-the-art approach. Full article

Computer vision technology is increasingly being used in areas such as intelligent security and autonomous driving. Users need accurate and reliable visual information, but the images obtained under severe weather conditions are often disturbed by rainy weather, causing image scenes to look blurry. Many current single image deraining algorithms achieve good performance but have limitations in retaining detailed image information. In this paper, we design a Scale-space Feature Recalibration Network (SFR-Net) for single image deraining. The proposed network improves the image feature extraction and characterization capability of a Multi-scale Extraction Recalibration Block (MERB) using dilated convolution with different convolution kernel sizes, which results in rich multi-scale rain streaks features. In addition, we develop a Subspace Coordinated Attention Mechanism (SCAM) and embed it into MERB, which combines coordinated attention recalibration and a subspace attention mechanism to recalibrate the rain streaks feature information learned from the feature extraction phase and eliminate redundant feature information to enhance the transfer of important feature information. Meanwhile, the overall SFR-Net structure uses dense connection and cross-layer feature fusion to repeatedly utilize the feature maps, thus enhancing the understanding of the network and avoiding gradient disappearance. Through extensive experiments on synthetic and real datasets, the proposed method outperforms the recent state-of-the-art deraining algorithms in terms of both the rain removal effect and the preservation of image detail information. Full article

In this paper, we focus on a challenging, but interesting, task in speech emotion recognition (SER), i.e., cross-corpus SER. Unlike conventional SER, a feature distribution mismatch may exist between the labeled source (training) and target (testing) speech samples in cross-corpus SER because they come from different speech emotion corpora, which degrades the performance of most well-performing SER methods. To address this issue, we propose a novel transfer subspace learning method called multiple distribution-adapted regression (MDAR) to bridge the gap between speech samples from different corpora. Specifically, MDAR aims to learn a projection matrix to build the relationship between the source speech features and emotion labels. A novel regularization term called multiple distribution adaption (MDA), consisting of a marginal and two conditional distribution-adapted operations, is designed to collaboratively enable such a discriminative projection matrix to be applicable to the target speech samples, regardless of speech corpus variance. Consequently, by resorting to the learned projection matrix, we are able to predict the emotion labels of target speech samples when only the source label information is given. To evaluate the proposed MDAR method, extensive cross-corpus SER tasks based on three different speech emotion corpora, i.e., EmoDB, eNTERFACE, and CASIA, were designed. Experimental results showed that the proposed MDAR outperformed most recent state-of-the-art transfer subspace learning methods and even performed better than several well-performing deep transfer learning methods in dealing with cross-corpus SER tasks. Full article

In this paper, we investigate the problem of cross-corpus speech emotion recognition (SER), in which the training (source) and testing (target) speech samples belong to different corpora. This case thus leads to a feature distribution mismatch between the source and target speech samples. Hence, the performance of most existing SER methods drops sharply. To solve this problem, we propose a simple yet effective transfer subspace learning method called joint distribution implicitly aligned subspace learning (JIASL). The basic idea of JIASL is very straightforward, i.e., building an emotion discriminative and corpus invariant linear regression model under an implicit distribution alignment strategy. Following this idea, we first make use of the source speech features and emotion labels to endow such a regression model with emotion-discriminative ability. Then, a well-designed reconstruction regularization term, jointly considering the marginal and conditional distribution alignments between the speech samples in both corpora, is adopted to implicitly enable the regression model to predict the emotion labels of target speech samples. To evaluate the performance of our proposed JIASL, extensive cross-corpus SER experiments are carried out, and the results demonstrate the promising performance of the proposed JIASL in coping with the tasks of cross-corpus SER. Full article

A pathological voice detection system is designed to detect pathological characteristics of vocal cords from speech. Such systems are particularly susceptible to domain mismatch where the training and testing data come from the source and target domains, respectively. Due to the difference in speech disease etiology, recording environment, and device, etc., the feature distributions of source and target domain are quite different. Meanwhile, considering the high costs of annotating labels, it is hard to acquire labeled data in the target domain. This paper attempts to formulate cross-domain pathological voice detection as an unsupervised domain adaptation problem. Joint subspace transfer learning (JSTL) aims to find a projection matrix to transform source and target domain data into a common space. The maximum mean discrepancy function is used to measure the divergence across databases. Intra-class and inter-class distance act as regularization to guarantee the maximum separability between different classes. A graph matrix is constructed to help transfer knowledge from the relevant source data to the target data. Three popular pathological voice databases were selected in this paper. For six cross-database experiments, the accuracy of the method proposed increased by up to 15%. For different voice categories, the category of structural voice showed the most significant increase, nearly 20%. Full article

Individual differences often appear in electroencephalography (EEG) data collected from different subjects due to its weak, nonstationary and low signal-to-noise ratio properties. This causes many machine learning methods to have poor generalization performance because the independent identically distributed assumption is no longer valid in cross-subject EEG data. To this end, transfer learning has been introduced to alleviate the data distribution difference between subjects. However, most of the existing methods have focused only on domain adaptation and failed to achieve effective collaboration with label estimation. In this paper, an EEG feature transfer method combined with semi-supervised regression and bipartite graph label propagation (TSRBG) is proposed to realize the unified joint optimization of EEG feature distribution alignment and semi-supervised joint label estimation. Through the cross-subject emotion recognition experiments on the SEED-IV data set, the results show that (1) TSRBG has significantly better recognition performance in comparison with the state-of-the-art models; (2) the EEG feature distribution differences between subjects are significantly minimized in the learned shared subspace, indicating the effectiveness of domain adaptation; (3) the key EEG frequency bands and channels for cross-subject EEG emotion recognition are achieved by investigating the learned subspace, which provides more insights into the study of EEG emotion activation patterns. Full article

In this work, an improved dynamic convolutional neural network (DCNN) model to accurately identify the risk of a forest fire was established based on the traditional DCNN model. First, the DCNN network model was trained in combination with transfer learning, and multiple pre-trained DCNN models were used to extract features from forest fire images. Second, principal component analysis (PCA) reconstruction technology was used in the appropriate subspace. The constructed 15-layer forest fire risk identification DCNN model named “DCN_Fire” could accurately identify core fire insurance areas. Moreover, the original and enhanced image data sets were used to evaluate the impact of data enhancement on the model’s accuracy. The traditional DCNN model was improved and the recognition speed and accuracy were compared and analyzed with the other three DCNN model algorithms with different architectures. The difficulty of using DCNN to monitor forest fire risk was solved, and the model’s detection accuracy was further improved. The true positive rate was 7.41% and the false positive rate was 4.8%. When verifying the impact of different batch sizes and loss rates on verification accuracy, the loss rate of the DCN_Fire model of 0.5 and the batch size of 50 provided the optimal value for verification accuracy (0.983). The analysis results showed that the improved DCNN model had excellent recognition speed and accuracy and could accurately recognize and classify the risk of a forest fire under natural light conditions, thereby providing a technical reference for preventing and tackling forest fires. Full article

Due to the vast geometric and radiometric differences between SAR and optical images, SAR-optical image matching remains an intractable challenge. Despite the fact that the deep learning-based matching model has achieved great success, SAR feature embedding ability is not fully explored yet because of the lack of well-designed pre-training techniques. In this paper, we propose to employ the self-supervised learning method in the SAR-optical matching framework, in order to serve as a pre-training strategy for improving the representation learning ability of SAR images as well as optical images. We first use a state-of-the-art self-supervised learning method, Momentum Contrast (MoCo), to pre-train an optical feature encoder and an SAR feature encoder separately. Then, the pre-trained encoders are transferred to an advanced common representation learning model, Bridge Neural Network (BNN), to project the SAR and optical images into a more distinguishable common feature representation subspace, which leads to a high multi-modal image matching result. Experimental results on three SAR-optical matching benchmark datasets show that our proposed MoCo pre-training method achieves a high matching accuracy up to 0.873 even for the complex QXS-SAROPT SAR-optical matching dataset. BNN pre-trained with MoCo outperforms BNN with the most commonly used ImageNet pre-training, and achieves at most 4.4% gains in matching accuracy. Full article

In recent years, few-shot remote sensing scene classification has attracted significant attention, aiming to obtain excellent performance under the condition of insufficient sample numbers. A few-shot remote sensing scene classification framework contains two phases: (i) the pre-training phase seeks to adopt base data to train a feature extractor, and (ii) the meta-testing phase uses the pre-training feature extractor to extract novel data features and design classifiers to complete classification tasks. Because of the difference in the data category, the pre-training feature extractor cannot adapt to the novel data category, named negative transfer problem. We propose a novel method for few-shot remote sensing scene classification based on shared class Sparse Principal Component Analysis (SparsePCA) to solve this problem. First, we propose, using self-supervised learning, to assist-train a feature extractor. We construct a self-supervised assisted classification task to improve the robustness of the feature extractor in the case of fewer training samples and make it more suitable for the downstream classification task. Then, we propose a novel classifier for the few-shot remote sensing scene classification named Class-Shared SparsePCA classifier (CSSPCA). The CSSPCA projects novel data features into subspace to make reconstructed features more discriminative and complete the classification task. We have conducted many experiments on remote sensing datasets, and the results show that the proposed method dramatically improves classification accuracy. Full article

Distribution discrepancies between different sessions greatly degenerate the performance of video-evoked electroencephalogram (EEG) emotion recognition. There are discrepancies since the EEG signal is weak and non-stationary and these discrepancies are manifested in different trails in each session and even in some trails which belong to the same emotion. To this end, we propose a Coupled Projection Transfer Metric Learning (CPTML) model to jointly complete domain alignment and graph-based metric learning, which is a unified framework to simultaneously minimize cross-session and cross-trial divergences. By experimenting on the SEED_IV emotional dataset, we show that (1) CPTML exhibits a significantly better performance than several other approaches; (2) the cross-session distribution discrepancies are minimized and emotion metric graph across different trials are optimized in the CPTML-induced subspace, indicating the effectiveness of data alignment and metric exploration; and (3) critical EEG frequency bands and channels for emotion recognition are automatically identified from the learned projection matrices, providing more insights into the occurrence of the effect. Full article