Query：
学者姓名：孔祥增
Refining：
Year
Type
Indexed by
Source
Complex
Co-Author
Language
Clean All
Abstract ：
The combination of electroencephalography (EEG) signals with peripheral physiological signals (PPS) for emotion recognition has garnered increasing attention due to their objectivity and rich information content. However, the heterogeneity and correlations of multimodal physiological signals remain significant challenges in multimodal emotion recognition. In addition, effectively extracting spatiotemporal features from these physiological signals is crucial for accurate emotion recognition. To address these issues, we propose a multimodal emotion recognition framework based on intra-and inter-modal supervised contrastive learning (IAECL), using EEG and PPS signals as inputs. Specifically, IAECL employs modality-specific encoders to mitigate heterogeneity. In addition, a dual-level supervised contrastive learning strategy is introduced: intra-modal contrastive learning enhances the feature compactness within each modality, and inter-modal contrastive learning strengthens cross-modal correlations. To better extract complex spatiotemporal features from EEG, we further a novel EEG encoder that integrates graph-based modeling, spatial-temporal convolution, and multi-head attention. This design effectively captures inter-channel dependencies and enhances local-global feature interactions to improve EEG feature representation. Extensive experiments on the publicly available DEAP and DREAMER datasets demonstrated the superior performance of IAECL, which achieved state-of-the-art results in both within-subject and mixed-subject emotion classification tasks. Visualization analyses further revealed that IAECL effectively aligns multimodal features to enhance inter-modal correlations, offering a novel solution to the challenges of multi-modal emotion recognition. We also analyzed the optimal combination of other physiological signals (e.g., electromyography or temperature recording signals) with EEG for multimodal emotion recognition.
Keyword ：
EEG EEG Emotion recognition Emotion recognition Multimodal physiological signal Multimodal physiological signal Spatiotemporal features Spatiotemporal features Supervised contrastive learning Supervised contrastive learning
Cite：
Copy from the list or Export to your reference management。
| GB/T 7714 | Zhao, Rongyue , Ma, Tianyuan , Li, Wangsen et al. EEG-based multimodal emotion recognition framework with supervised contrastive learning and spatial-temporal convolutional attention network [J]. | DIGITAL SIGNAL PROCESSING , 2026 , 168 . |
| MLA | Zhao, Rongyue et al. "EEG-based multimodal emotion recognition framework with supervised contrastive learning and spatial-temporal convolutional attention network" . | DIGITAL SIGNAL PROCESSING 168 (2026) . |
| APA | Zhao, Rongyue , Ma, Tianyuan , Li, Wangsen , Li, Yihua , Li, Jinghu , Huang, Dong et al. EEG-based multimodal emotion recognition framework with supervised contrastive learning and spatial-temporal convolutional attention network . | DIGITAL SIGNAL PROCESSING , 2026 , 168 . |
| Export to | NoteExpress RIS BibTex |
Version ：
Abstract ：
Decoding brain activity from non-invasive motor imagery electroencephalography (MI-EEG) signals is vital for brain-computer interface (BCI). Many studies overlook the feature-classifier interaction, impacting decoding performance. To tackle this issue, we propose a decoding framework for MI-EEG signals that uses multi-view weighted features. Our approach begins by employing a multi-view feature fusion mechanism to capture both global and local features from raw MI-EEG signals. Following feature extraction and fusion, we utilize the Expectation-Maximization (EM) algorithm to partition the samples to distinct soft subspaces. The subset information is subsequently used to classify uncategorized MI-EEG samples. Within this framework, features that are more relevant to the subsequent classification model are assigned greater weights. Results from public benchmark datasets demonstrate that, compared to commonly used models, our method achieves superior classification results. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.
Keyword ：
Benchmarking Benchmarking Biomedical signal processing Biomedical signal processing Brain mapping Brain mapping Electroencephalography Electroencephalography Expectation maximization algorithm Expectation maximization algorithm
Cite：
Copy from the list or Export to your reference management。
| GB/T 7714 | Li, Nan , Li, Wangsen , Zhang, Tingting et al. Motor Imagery EEG Signals Decoding with Multi-view Weighted Features [C] . 2025 : 338-349 . |
| MLA | Li, Nan et al. "Motor Imagery EEG Signals Decoding with Multi-view Weighted Features" . (2025) : 338-349 . |
| APA | Li, Nan , Li, Wangsen , Zhang, Tingting , Huang, Dong , Han, Junfeng , Kong, Xiangzeng . Motor Imagery EEG Signals Decoding with Multi-view Weighted Features . (2025) : 338-349 . |
| Export to | NoteExpress RIS BibTex |
Version ：
Abstract ：
In recent years, food fraud issues related to whey protein supplements have disrupted the market and caused significant concern among consumers. Conventional analytical methods such as HPLC and ion exchange chromatography are commonly used to detect adulteration in whey protein supplements. However, these methods are costly, time-consuming and require specialised operation, making them less suitable for a wider range of users. This study presents a rapid and reliable approach for verifying the authenticity of whey protein supplements using laser-induced breakdown spectroscopy (LIBS) and machine learning. Specifically, this approach is employed to identify 15 brands of whey protein concentration (WPC), quantify protein and carbohydrate concentrations, distinguish three types of adulterants, and predict the level of adulteration. The relationship between LIBS data and analyte labels is established using machine learning methods, including partial least squares regression (PLSR), partial least squares discriminant analysis (PLS-DA), and kernel extreme learning machine (K-ELM). The accuracy for identifying WPC brands was over 0.977, and the highest coefficient of determination (R2) for quantifying protein and carbohydrate contents was 0.984 and 0.978, respectively. In addition, different adulterants can be differentiated with accuracies exceeding 0.986, and the R2 values for adulteration prediction are above 0.967 in most cases. These results suggest that LIBS combined with machine learning can serve as a viable and efficient solution for detecting adulteration in whey protein supplements.
Keyword ：
authentication authentication kernel extreme learning machine kernel extreme learning machine laser-induced breakdown spectroscopy laser-induced breakdown spectroscopy machine learning machine learning partial least squares partial least squares Whey protein concentrate Whey protein concentrate
Cite：
Copy from the list or Export to your reference management。
| GB/T 7714 | Zhu, Meiling , Song, Weiran , Tang, Xuan et al. Detection of whey protein concentrate adulteration using laser-induced breakdown spectroscopy combined with machine learning [J]. | FOOD ADDITIVES AND CONTAMINANTS PART A-CHEMISTRY ANALYSIS CONTROL EXPOSURE & RISK ASSESSMENT , 2025 , 42 (5) : 570-579 . |
| MLA | Zhu, Meiling et al. "Detection of whey protein concentrate adulteration using laser-induced breakdown spectroscopy combined with machine learning" . | FOOD ADDITIVES AND CONTAMINANTS PART A-CHEMISTRY ANALYSIS CONTROL EXPOSURE & RISK ASSESSMENT 42 . 5 (2025) : 570-579 . |
| APA | Zhu, Meiling , Song, Weiran , Tang, Xuan , Kong, Xiangzeng . Detection of whey protein concentrate adulteration using laser-induced breakdown spectroscopy combined with machine learning . | FOOD ADDITIVES AND CONTAMINANTS PART A-CHEMISTRY ANALYSIS CONTROL EXPOSURE & RISK ASSESSMENT , 2025 , 42 (5) , 570-579 . |
| Export to | NoteExpress RIS BibTex |
Version ：
Abstract ：
With the growing interest in health and fitness, whey protein supplements are becoming increasingly popular among fitness enthusiasts and athletes. The surge in demand for whey protein supplements highlights the need for cost-effective methods to characterise product quality throughout the food supply chain. This study presents a rapid and low-cost method for authenticating sports whey protein supplements using smartphone video imaging (SVI) combined with machine learning. A gradient of colours ranging from purple to red is displayed on the front screen of a smartphone to illuminate the sample. The colour change on the sample surface is captured in a short video by the front-facing camera. Then, the video is split into frames, decomposed into RGB colour channels, and converted into spectral data. The relationship between video data and sample labels is established using machine learning models. The proposed method is tested on five tasks, including identifying 15 brands of whey protein concentrate (WPC), quantifying fat content and energy levels, detecting three types of adulterants, and quantifying adulterant levels. Moreover, the performance of SVI was compared to that of hyperspectral imaging (HSI), which has an equipment cost of around 80 times that of SVI. The proposed method achieves accuracies of 0.933 and 0.96 in WPC brand identification and adulterant detection, respectively, which are only around 0.05 lower than those of HSI. It obtains coefficients of determination of 0.897, 0.906 and 0.963 for the quantification of fat content, energy levels and milk powder adulteration, respectively. Such results demonstrate that the combination of smartphones and machine learning offers a low-cost and viable preliminary screening tool for verifying the authenticity of whey protein supplements.
Keyword ：
authentication authentication chemometrics chemometrics machine learning machine learning smartphone video imaging smartphone video imaging whey protein concentrate whey protein concentrate
Cite：
Copy from the list or Export to your reference management。
| GB/T 7714 | Tang, Xuan , Du, Wenjiao , Song, Weiran et al. Smartphone Video Imaging Combined with Machine Learning: A Cost-Effective Method for Authenticating Whey Protein Supplements [J]. | FOODS , 2025 , 14 (7) . |
| MLA | Tang, Xuan et al. "Smartphone Video Imaging Combined with Machine Learning: A Cost-Effective Method for Authenticating Whey Protein Supplements" . | FOODS 14 . 7 (2025) . |
| APA | Tang, Xuan , Du, Wenjiao , Song, Weiran , Gu, Weilun , Kong, Xiangzeng . Smartphone Video Imaging Combined with Machine Learning: A Cost-Effective Method for Authenticating Whey Protein Supplements . | FOODS , 2025 , 14 (7) . |
| Export to | NoteExpress RIS BibTex |
Version ：
Abstract ：
Background: The residual fluoroquinolone antibiotics (FQs) in the environment and food has raised public concerns over their potential impact on human health. Terahertz metamaterial sensors (TMSs) have garnered significant attention due to their capability to enhance the interaction between terahertz waves and antibiotic molecules, enabling the detection of trace antibiotics. However, conventional quantitative and qualitative methods based on TMSs suffer from low accuracy and cumbersome processes, respectively. Herein, this work proposed a novel approach that reconstructed optimal terahertz response features of different TMSs with machine learning algorithms, which allowed for analysis of three similar trace FQs with enhanced accuracy. Results: The prepared three patterned TMSs exhibited different resonance responses, which varied with changes in FQs types and concentrations. The resonance peak features of the three TMSs were fused to construct the resonance peak feature matrix (W0) and combined with the K-Nearest Neighbor (KNN) algorithm to build the W0-KNN classification model. The interval feature matrix was constructed by optimizing and expanding the resonance peak feature width. The optimal resonance peak interval feature matrix (Wt) was combined with Gaussian process regression (GPR) algorithms with different kernel functions to build the Wt-GPR quantitative model. The results showed that W0-KNN achieved 100 % classification accuracy for the three FQs. Wt-GPR exhibited high quantitative accuracy for all three FQs with the determination coefficient (R2) of 0.94-0.98, and root mean square error (RMSE) of 6.4085-10.6540. The results of Wt-GPR with different kernel functions had small fluctuations, demonstrating high stability in predictive performance. Significance: Reconstructing features from multi-TMSs in combination with machine learning algorithms enables rapid, precise, and reliable qualitative and quantitative analysis of trace FQs. Our research introduces innovative concepts and methodologies to detect trace FQs using TMS-based sensors, paving the way for future applications of TMS in the biomolecular sensing and detection.
Keyword ：
Biomolecular sensing Biomolecular sensing Fluoroquinolone antibiotics Fluoroquinolone antibiotics Machine learning Machine learning Terahertz metamaterials sensor Terahertz metamaterials sensor Trace detection Trace detection
Cite：
Copy from the list or Export to your reference management。
| GB/T 7714 | Zhang, Lintong , Kong, Xiangzeng , Wang, Shuhui et al. Resonance features integration of multiple terahertz metamaterials sensors for qualification and quantification of trace fluoroquinolone antibiotics [J]. | ANALYTICA CHIMICA ACTA , 2025 , 1345 . |
| MLA | Zhang, Lintong et al. "Resonance features integration of multiple terahertz metamaterials sensors for qualification and quantification of trace fluoroquinolone antibiotics" . | ANALYTICA CHIMICA ACTA 1345 (2025) . |
| APA | Zhang, Lintong , Kong, Xiangzeng , Wang, Shuhui , Zhang, Wenqing , Wu, Libin , Liu, Xinze et al. Resonance features integration of multiple terahertz metamaterials sensors for qualification and quantification of trace fluoroquinolone antibiotics . | ANALYTICA CHIMICA ACTA , 2025 , 1345 . |
| Export to | NoteExpress RIS BibTex |
Version ：
Abstract ：
Electroencephalography (EEG) serves as a practical auxiliary tool deployed to diagnose diverse brain-related disorders owing to its exceptional temporal resolution, non-invasive characteristics, and cost-effectiveness. In recent years, with the advancement of machine learning, automated EEG pathology diagnostics methods have flourished. However, most existing methods usually neglect the crucial spatial correlations in multi-channel EEG signals and the potential complementary information among different domain features, both of which are keys to improving discrimination performance. In addition, latent redundant and irrelevant features may cause overfitting, increased model complexity, and other issues. In response, we propose a novel feature-based framework designed to improve the diagnostic accuracy of multi-channel EEG pathology. This framework first applies a multi-resolution decomposition technique and a statistical feature extractor to construct a salient time-frequency feature space. Then, spatial distribution information is channel-wise extracted from this space to fuse with time-frequency features, thereby leveraging their complementarity to the fullest extent. Furthermore, to eliminate the redundancy and irrelevancy, a two-step dimension reduction strategy, including a lightweight multi-view time-frequency feature aggregation and a non-parametric statistical significance analysis, is devised to pick out the features with stronger discriminative ability. Comprehensive examinations of the Temple University Hospital Abnormal EEG Corpus V. 2.0.0 demonstrate that our proposal outperforms state-of-the-art methods, highlighting its significant potential in clinically automated EEG abnormality detection.
Cite：
Copy from the list or Export to your reference management。
| GB/T 7714 | Chen, Shimiao , Huang, Dong , Liu, Xinyue et al. Automatic diagnostics of electroencephalography pathology based on multi-domain feature fusion [J]. | PLOS ONE , 2025 , 20 (5) . |
| MLA | Chen, Shimiao et al. "Automatic diagnostics of electroencephalography pathology based on multi-domain feature fusion" . | PLOS ONE 20 . 5 (2025) . |
| APA | Chen, Shimiao , Huang, Dong , Liu, Xinyue , Chen, Jianjun , Kong, Xiangzeng , Zhang, Tingting . Automatic diagnostics of electroencephalography pathology based on multi-domain feature fusion . | PLOS ONE , 2025 , 20 (5) . |
| Export to | NoteExpress RIS BibTex |
Version ：
Abstract ：
Near-infrared (NIR) spectroscopy, with its advantages of non-destructive analysis, simple operation, and fast detection speed, has been widely applied in various fields. However, the effectiveness of current spectral analysis techniques still relies on complex preprocessing and feature selection of spectral data. While data-driven deep learning can automatically extract features from raw spectral data, it typically requires large amounts of labeled data for training, limiting its application in spectral analysis. To address this issue, we propose a self-supervised learning (SSL) framework based on convolutional neural networks (CNN) to enhance spectral analysis performance with small sample sizes. The method comprises two learning stages: pre-training and fine-tuning. In the pre-training stage, a large amount of pseudo-labeled data is used to learn intrinsic spectral features, followed by fine-tuning with a smaller set of labeled data to complete the final model training. Applied to our own collected dataset of three tea varieties, the proposed model achieved a classification accuracy of 99.12%. Additionally, experiments on three public datasets demonstrated that the SSL model significantly outperforms traditional machine learning methods, achieving accuracies of 97.83%, 98.14%, and 99.89%, respectively. Comparative experiments further confirmed the effectiveness of the pre-training stage, with the highest accuracy improvement, reaching 10.41%. These results highlight the potential of the proposed method for handling small sample spectral data, providing a viable solution for improved spectral analysis.
Cite：
Copy from the list or Export to your reference management。
| GB/T 7714 | Zhao, Rongyue , Li, Wangsen , Xu, Jinchai et al. A CNN-based self-supervised learning framework for small-sample near-infrared spectroscopy classification [J]. | ANALYTICAL METHODS , 2025 , 17 (5) : 1090-1100 . |
| MLA | Zhao, Rongyue et al. "A CNN-based self-supervised learning framework for small-sample near-infrared spectroscopy classification" . | ANALYTICAL METHODS 17 . 5 (2025) : 1090-1100 . |
| APA | Zhao, Rongyue , Li, Wangsen , Xu, Jinchai , Chen, Linjie , Wei, Xuan , Kong, Xiangzeng . A CNN-based self-supervised learning framework for small-sample near-infrared spectroscopy classification . | ANALYTICAL METHODS , 2025 , 17 (5) , 1090-1100 . |
| Export to | NoteExpress RIS BibTex |
Version ：
Abstract ：
Bird species monitoring is crucial for conservation, but overlapping vocalizations in natural environments complicate multi-label classification, affecting model performance. To address these issues, we developed the Adaptive Multi-label Attention Threshold Network (AMAT-Net) as a bird sound classification framework. AMAT-Net employs a multi-view strategy, combining bidirectional gated recurrent unit (BiGRU)-attention networks analyze temporal features and multi-scale convolutional neural networks to extract spectral features, enabling analysis of bird sounds. Given the differences between temporal and spectral features, time-domain features capture transient changes, whereas frequency-domain features reveal spectral trends. Balancing the essential features of both without losing details is difficult. Therefore, we designed the temporal-spectral attention feature fusion (TSAFF) module to optimize feature fusion. TSAFF employs an attention-based mechanism fuse temporal and spectral features, enhancing the cross-domain feature complementarity. Binary classification is conducted between relevant and irrelevant labels, and threshold is determined based on the results. A score based thresholding strategy called dynamic threshold scaling was then developed. A label correlation matrix constructed using Pearson's correlation coefficients, and the classifier's scores for instance-label pairs with high inter-label correlations are adjusted accordingly during prediction. In addition, hierarchical cross-validation used to search for the threshold that maximizes the F1 score, dynamically optimizing the decision boundary for each species to adapt to the actual label distribution. Experimental results on a synthesized dataset of 10 bird species (including cases of 2, 3, and 4 species vocalizing simultaneously) and the public BirdCLEF+2025 dataset demonstrate that AMAT-Net achieves an accuracy of 95.54% with a macro-F1 score of 91.26% on the synthesized dataset, and an accuracy of 98.75% with a macro-F1 score of 93.14% on BirdCLEF+2025.
Keyword ：
Dynamic threshold Dynamic threshold Feature fusion Feature fusion Label correlation Label correlation Multi-label classification Multi-label classification Multi-view Multi-view
Cite：
Copy from the list or Export to your reference management。
| GB/T 7714 | Fang, Minghui , Wu, Dengwei , Wu, Wei et al. Multi-label bird sound recognition based on multi-view learning and dynamic threshold adjustment [J]. | APPLIED ACOUSTICS , 2025 , 240 . |
| MLA | Fang, Minghui et al. "Multi-label bird sound recognition based on multi-view learning and dynamic threshold adjustment" . | APPLIED ACOUSTICS 240 (2025) . |
| APA | Fang, Minghui , Wu, Dengwei , Wu, Wei , Fang, Mengfan , Chen, Yanhong , Kong, Xiangzeng et al. Multi-label bird sound recognition based on multi-view learning and dynamic threshold adjustment . | APPLIED ACOUSTICS , 2025 , 240 . |
| Export to | NoteExpress RIS BibTex |
Version ：
Abstract ：
Flexible strain sensors have garnered significant attention due to their high sensitivity, rapid response, and flexibility. Recent innovations, particularly those incorporating machine learning, have significantly enhanced their stability, sensitivity, and adaptability, positioning these sensors as promising solutions in health monitoring, human-computer interaction, and smart home applications. However, challenges remain in optimizing sensor materials for enhanced responsiveness, durability, and stability. Moreover, the development of machine learning-based strain sensors faces obstacles, including algorithmic limitations, low noise tolerance in complex environments, and limited model interpretability. This review systematically evaluates the latest advancements in flexible strain sensors, emphasizing the critical role of machine learning in performance enhancement. It further explores the shift from traditional machine learning methods to deep learning approaches, elucidating the potential applications that these algorithms facilitate. Finally, we discuss future research trajectories, highlighting both opportunities and challenges that may guide the next wave of innovations in this dynamic field.
Keyword ：
deeplearning deeplearning flexible electronics flexible electronics machine learning machine learning sensing material sensing material smart strain sensor smart strain sensor
Cite：
Copy from the list or Export to your reference management。
| GB/T 7714 | Kong, Xiangzeng , Wen, Wangxiao , Guan, Yujie et al. Advances in Machine Learning-Driven Flexible Strain Sensors: Challenges, Innovations, and Applications [J]. | ACS APPLIED MATERIALS & INTERFACES , 2025 , 17 (22) : 31778-31798 . |
| MLA | Kong, Xiangzeng et al. "Advances in Machine Learning-Driven Flexible Strain Sensors: Challenges, Innovations, and Applications" . | ACS APPLIED MATERIALS & INTERFACES 17 . 22 (2025) : 31778-31798 . |
| APA | Kong, Xiangzeng , Wen, Wangxiao , Guan, Yujie , Lin, Zihan , Zheng, Junwei , Xie, Banghao et al. Advances in Machine Learning-Driven Flexible Strain Sensors: Challenges, Innovations, and Applications . | ACS APPLIED MATERIALS & INTERFACES , 2025 , 17 (22) , 31778-31798 . |
| Export to | NoteExpress RIS BibTex |
Version ：
Abstract ：
The residues of fluoroquinolone antibiotics (FQs) have attracted widespread attention due to their potential health risks. Given the high structural similarity among FQs, traditional methods for analyzing their mixed residues are often complex and inaccurate. This study proposed a method combining terahertz (THz) spectroscopy and machine learning for qualitative and quantitative analysis of multicomponent mixtures containing four FQs (nadifloxacin [NAD], pefloxacin [PEF], ofloxacin [OFL], and enrofloxacin [ENR]) at low mass ratios (0.067-0.333). For qualitative analysis, multi-step preprocessing (MP) and support vector machine (SVM) were integrated to develop the MP-SVM model. The MP-SVM achieved an average classification accuracy of 0.967 for the four FQs in test sets. While MP enabled high-resolution qualitative discrimination, it was insufficient for accurate quantification of complex FQ mixtures. Consequently, further optimizations of feature data and model parameters were conducted for quantitative analysis. Specifically, a high-quality feature matrix (T) was constructed by merging fingerprint features of each FQ. The sparrow search algorithm (SSA) was employed to optimize support vector regression (SVR) parameters, forming the MP-T-SSA-SVR model. This model significantly improved quantitative performance, with a coefficient of determination (R2) of 0.971-0.972, a root mean square error (RMSE) of 3.405-3.514, and a mean absolute error (MAE) of 2.090-2.400 in test sets. Compared to similar studies, this work involves more diverse FQs with higher structural similarity, providing a new reference for advancing qualitative and quantitative analysis of practical multicomponent mixtures.
Keyword ：
Fluoroquinolone antibiotics Fluoroquinolone antibiotics Multicomponent mixtures Multicomponent mixtures Qualitative and quantitative analysis Qualitative and quantitative analysis Similar molecular structures Similar molecular structures Terahertz spectroscopy Terahertz spectroscopy
Cite：
Copy from the list or Export to your reference management。
| GB/T 7714 | Zhang, Lintong , Liu, Xinze , Yang, Jingsen et al. Qualitative and Quantitative Analysis of Multivariate Mixed Fluoroquinolone Antibiotics Based on Terahertz Spectroscopy [J]. | JOURNAL OF INFRARED MILLIMETER AND TERAHERTZ WAVES , 2025 , 46 (10) . |
| MLA | Zhang, Lintong et al. "Qualitative and Quantitative Analysis of Multivariate Mixed Fluoroquinolone Antibiotics Based on Terahertz Spectroscopy" . | JOURNAL OF INFRARED MILLIMETER AND TERAHERTZ WAVES 46 . 10 (2025) . |
| APA | Zhang, Lintong , Liu, Xinze , Yang, Jingsen , Wang, Shuhui , Zhang, Jiachen , Yang, Wangjincheng et al. Qualitative and Quantitative Analysis of Multivariate Mixed Fluoroquinolone Antibiotics Based on Terahertz Spectroscopy . | JOURNAL OF INFRARED MILLIMETER AND TERAHERTZ WAVES , 2025 , 46 (10) . |
| Export to | NoteExpress RIS BibTex |
Version ：
Export
| Results： |
Selected to |
| Format： |
