Underwater monopulse space-time adaptive track-before-detect method,which combines space-time adaptive detector(STAD)and the track-before-detect algorithm based on dynamic programming(DP-TBD),denoted as STAD-DP-TBD,ca...Underwater monopulse space-time adaptive track-before-detect method,which combines space-time adaptive detector(STAD)and the track-before-detect algorithm based on dynamic programming(DP-TBD),denoted as STAD-DP-TBD,can effectively detect low-speed weak targets.However,due to the complexity and variability of the underwater environment,it is difficult to obtain sufficient secondary data,resulting in a serious decline in the detection and tracking performance,and leading to poor robustness of the algorithm.In this paper,based on the adaptive matched filter(AMF)test and the RAO test,underwater monopulse AMF-DP-TBD algorithm and RAO-DP-TBD algorithm which incorporate persymmetry and symmetric spectrum,denoted as PSAMF-DP-TBD and PS-RAO-DP-TBD,are proposed and compared with the AMF-DP-TBD algorithm and RAO-DP-TBD algorithm based on persymmetry array,denoted as P-AMF-DP-TBD and P-RAO-DP-TBD.The simulation results show that the four methods can work normally with sufficient secondary data and slightly insufficient secondary data,but when the secondary data is severely insufficient,the P-AMF-DP-TBD and P-RAO-DP-TBD algorithms has failed while the PSAMF-DP-TBD and PS-RAO-DP-TBD algorithms still have good detection and tracking capabilities.展开更多
Moderate exercise contributes to health, but excessive exercise may lead to physicalinjury or even endanger life. It is pressing for a device that can detect the intensity of exercise.Therefore, in order to enable rea...Moderate exercise contributes to health, but excessive exercise may lead to physicalinjury or even endanger life. It is pressing for a device that can detect the intensity of exercise.Therefore, in order to enable real-time detection of exercise intensity and mitigate the risks of harmfrom excessive exercise, a exercise intensity monitoring system based on the heart rate variability(HRV) from electrocardiogram (ECG) signal and linear features from phonocardiogram (PCG)signal is proposed. The main contributions include: First, accurate analysis of HRV is crucial forsubsequent exercise intensity detection. To enhance HRV analysis, we propose an R-peak detectorbased on encoder-decoder and temporal convolutional network (TCN). Experimental resultsdemonstrate that the proposed R-peak detector achieves an F1 score exceeding 0.99 on real high-intensity exercise ECG datasets. Second, an exercise fatigue monitoring system based on multi-signal feature fusion is proposed. Initially, utilizing the proposed R-peak detector for HRV extractionin exercise intensity detection,which outperforms traditional algorithms, with the system achieving a classification performance of 0.933 sensitivity, 0.802 specificity, and 0.960 accuracy. To further improve the system, we combine HRV with the linear features of PCG. Our exercise intensitydetection system achieves 90.2% specificity, 96.7% recall, and 98.1% accuracy in five-fold cross-validation.展开更多
Underwater direction of arrival(DOA)estimation has always been a very challenging theoretical and practical problem.Due to the serious non-stationary,non-linear,and non-Gaussian characteristics,machine learning based ...Underwater direction of arrival(DOA)estimation has always been a very challenging theoretical and practical problem.Due to the serious non-stationary,non-linear,and non-Gaussian characteristics,machine learning based DOA estimation methods trained on simulated Gaussian noised array data cannot be directly applied to actual underwater DOA estimation tasks.In order to deal with this problem,environmental data with no target echoes can be employed to analyze the non-Gaussian components.Then,the obtained information about non-Gaussian components can be used to whiten the array data.Based on these considerations,a novel practical sonar array whitening method was proposed.Specifically,based on a weak assumption that the non-Gaussian components in adjacent patches with and without target echoes are almost the same,canonical cor-relation analysis(CCA)and non-negative matrix factorization(NMF)techniques are employed for whitening the array data.With the whitened array data,machine learning based DOA estimation models trained on simulated Gaussian noised datasets can be used to perform underwater DOA estimation tasks.Experimental results illustrated that,using actual underwater datasets for testing with known machine learning based DOA estimation models,accurate and robust DOA estimation performance can be achieved by using the proposed whitening method in different underwater con-ditions.展开更多
Three dimensional(3-D)imaging algorithms with irregular planar multiple-input-multiple-output(MIMO)arrays are discussed and compared with each other.Based on the same MIMO array,a modified back projection algorithm(MB...Three dimensional(3-D)imaging algorithms with irregular planar multiple-input-multiple-output(MIMO)arrays are discussed and compared with each other.Based on the same MIMO array,a modified back projection algorithm(MBPA)is accordingly proposed and four imaging algorithms are used for comparison,back-projection method(BP),back-projection one in time domain(BP-TD),modified back-projection one and fast Fourier transform(FFT)-based MIMO range migration algorithm(FFT-based MIMO RMA).All of the algorithms have been implemented in practical application scenarios by use of the proposed imaging system.Back to the practical applications,MIMO array-based imaging system with wide-bandwidth properties provides an efficient tool to detect objects hidden behind a wall.An MIMO imaging radar system,composed of a vector network analyzer(VNA),a set of switches,and an array of Vivaldi antennas,have been designed,fabricated,and tested.Then,these algorithms have been applied to measured data collected in different scenarios constituted by five metallic spheres in the absence and in the presence of a wall between the antennas and the targets in simulation and pliers in free space for experimental test.Finally,the focusing properties and time consumption of the above algorithms are compared.展开更多
Handheld ultrasound devices are known for their portability and affordability,making them widely utilized in underdeveloped areas and community healthcare for rapid diagnosis and early screening.However,the image qual...Handheld ultrasound devices are known for their portability and affordability,making them widely utilized in underdeveloped areas and community healthcare for rapid diagnosis and early screening.However,the image quality of handheld ultrasound devices is not always satisfactory due to the limited equipment size,which hinders accurate diagnoses by doctors.At the same time,paired ultrasound images are difficult to obtain from the clinic because imaging process is complicated.Therefore,we propose a modified cycle generative adversarial network(cycleGAN) for ultrasound image enhancement from multiple organs via unpaired pre-training.We introduce an ultrasound image pre-training method that does not require paired images,alleviating the requirement for large-scale paired datasets.We also propose an enhanced block with different structures in the pre-training and fine-tuning phases,which can help achieve the goals of different training phases.To improve the robustness of the model,we add Gaussian noise to the training images as data augmentation.Our approach is effective in obtaining the best quantitative evaluation results using a small number of parameters and less training costs to improve the quality of handheld ultrasound devices.展开更多
Hyper-and multi-spectral image fusion is an important technology to produce hyper-spectral and hyper-resolution images,which always depends on the spectral response function andthe point spread function.However,few wo...Hyper-and multi-spectral image fusion is an important technology to produce hyper-spectral and hyper-resolution images,which always depends on the spectral response function andthe point spread function.However,few works have been payed on the estimation of the two degra-dation functions.To learn the two functions from image pairs to be fused,we propose a Dirichletnetwork,where both functions are properly constrained.Specifically,the spatial response function isconstrained with positivity,while the Dirichlet distribution along with a total variation is imposedon the point spread function.To the best of our knowledge,the neural network and the Dirichlet regularization are exclusively investigated,for the first time,to estimate the degradation functions.Both image degradation and fusion experiments demonstrate the effectiveness and superiority of theproposed Dirichlet network.展开更多
Human posture estimation is a prominent research topic in the fields of human-com-puter interaction,motion recognition,and other intelligent applications.However,achieving highaccuracy in key point localization,which ...Human posture estimation is a prominent research topic in the fields of human-com-puter interaction,motion recognition,and other intelligent applications.However,achieving highaccuracy in key point localization,which is crucial for intelligent applications,contradicts the lowdetection accuracy of human posture detection models in practical scenarios.To address this issue,a human pose estimation network called AT-HRNet has been proposed,which combines convolu-tional self-attention and cross-dimensional feature transformation.AT-HRNet captures significantfeature information from various regions in an adaptive manner,aggregating them through convolu-tional operations within the local receptive domain.The residual structures TripNeck and Trip-Block of the high-resolution network are designed to further refine the key point locations,wherethe attention weight is adjusted by a cross-dimensional interaction to obtain more features.To vali-date the effectiveness of this network,AT-HRNet was evaluated using the COCO2017 dataset.Theresults show that AT-HRNet outperforms HRNet by improving 3.2%in mAP,4.0%in AP75,and3.9%in AP^(M).This suggests that AT-HRNet can offer more beneficial solutions for human posture estimation.展开更多
基金supported by the National Natural Science Foundation of China (No.61971412)。
文摘Underwater monopulse space-time adaptive track-before-detect method,which combines space-time adaptive detector(STAD)and the track-before-detect algorithm based on dynamic programming(DP-TBD),denoted as STAD-DP-TBD,can effectively detect low-speed weak targets.However,due to the complexity and variability of the underwater environment,it is difficult to obtain sufficient secondary data,resulting in a serious decline in the detection and tracking performance,and leading to poor robustness of the algorithm.In this paper,based on the adaptive matched filter(AMF)test and the RAO test,underwater monopulse AMF-DP-TBD algorithm and RAO-DP-TBD algorithm which incorporate persymmetry and symmetric spectrum,denoted as PSAMF-DP-TBD and PS-RAO-DP-TBD,are proposed and compared with the AMF-DP-TBD algorithm and RAO-DP-TBD algorithm based on persymmetry array,denoted as P-AMF-DP-TBD and P-RAO-DP-TBD.The simulation results show that the four methods can work normally with sufficient secondary data and slightly insufficient secondary data,but when the secondary data is severely insufficient,the P-AMF-DP-TBD and P-RAO-DP-TBD algorithms has failed while the PSAMF-DP-TBD and PS-RAO-DP-TBD algorithms still have good detection and tracking capabilities.
基金the National Natural Science Foundation of China(No.62301056)the Fundamental Research Funds for Central Universities(No.2022QN005).
文摘Moderate exercise contributes to health, but excessive exercise may lead to physicalinjury or even endanger life. It is pressing for a device that can detect the intensity of exercise.Therefore, in order to enable real-time detection of exercise intensity and mitigate the risks of harmfrom excessive exercise, a exercise intensity monitoring system based on the heart rate variability(HRV) from electrocardiogram (ECG) signal and linear features from phonocardiogram (PCG)signal is proposed. The main contributions include: First, accurate analysis of HRV is crucial forsubsequent exercise intensity detection. To enhance HRV analysis, we propose an R-peak detectorbased on encoder-decoder and temporal convolutional network (TCN). Experimental resultsdemonstrate that the proposed R-peak detector achieves an F1 score exceeding 0.99 on real high-intensity exercise ECG datasets. Second, an exercise fatigue monitoring system based on multi-signal feature fusion is proposed. Initially, utilizing the proposed R-peak detector for HRV extractionin exercise intensity detection,which outperforms traditional algorithms, with the system achieving a classification performance of 0.933 sensitivity, 0.802 specificity, and 0.960 accuracy. To further improve the system, we combine HRV with the linear features of PCG. Our exercise intensitydetection system achieves 90.2% specificity, 96.7% recall, and 98.1% accuracy in five-fold cross-validation.
基金supported by the National Natural Science Foundation of China(No.51279033).
文摘Underwater direction of arrival(DOA)estimation has always been a very challenging theoretical and practical problem.Due to the serious non-stationary,non-linear,and non-Gaussian characteristics,machine learning based DOA estimation methods trained on simulated Gaussian noised array data cannot be directly applied to actual underwater DOA estimation tasks.In order to deal with this problem,environmental data with no target echoes can be employed to analyze the non-Gaussian components.Then,the obtained information about non-Gaussian components can be used to whiten the array data.Based on these considerations,a novel practical sonar array whitening method was proposed.Specifically,based on a weak assumption that the non-Gaussian components in adjacent patches with and without target echoes are almost the same,canonical cor-relation analysis(CCA)and non-negative matrix factorization(NMF)techniques are employed for whitening the array data.With the whitened array data,machine learning based DOA estimation models trained on simulated Gaussian noised datasets can be used to perform underwater DOA estimation tasks.Experimental results illustrated that,using actual underwater datasets for testing with known machine learning based DOA estimation models,accurate and robust DOA estimation performance can be achieved by using the proposed whitening method in different underwater con-ditions.
基金National Natural Science Foundation of China(No.62293493)。
文摘Three dimensional(3-D)imaging algorithms with irregular planar multiple-input-multiple-output(MIMO)arrays are discussed and compared with each other.Based on the same MIMO array,a modified back projection algorithm(MBPA)is accordingly proposed and four imaging algorithms are used for comparison,back-projection method(BP),back-projection one in time domain(BP-TD),modified back-projection one and fast Fourier transform(FFT)-based MIMO range migration algorithm(FFT-based MIMO RMA).All of the algorithms have been implemented in practical application scenarios by use of the proposed imaging system.Back to the practical applications,MIMO array-based imaging system with wide-bandwidth properties provides an efficient tool to detect objects hidden behind a wall.An MIMO imaging radar system,composed of a vector network analyzer(VNA),a set of switches,and an array of Vivaldi antennas,have been designed,fabricated,and tested.Then,these algorithms have been applied to measured data collected in different scenarios constituted by five metallic spheres in the absence and in the presence of a wall between the antennas and the targets in simulation and pliers in free space for experimental test.Finally,the focusing properties and time consumption of the above algorithms are compared.
文摘Handheld ultrasound devices are known for their portability and affordability,making them widely utilized in underdeveloped areas and community healthcare for rapid diagnosis and early screening.However,the image quality of handheld ultrasound devices is not always satisfactory due to the limited equipment size,which hinders accurate diagnoses by doctors.At the same time,paired ultrasound images are difficult to obtain from the clinic because imaging process is complicated.Therefore,we propose a modified cycle generative adversarial network(cycleGAN) for ultrasound image enhancement from multiple organs via unpaired pre-training.We introduce an ultrasound image pre-training method that does not require paired images,alleviating the requirement for large-scale paired datasets.We also propose an enhanced block with different structures in the pre-training and fine-tuning phases,which can help achieve the goals of different training phases.To improve the robustness of the model,we add Gaussian noise to the training images as data augmentation.Our approach is effective in obtaining the best quantitative evaluation results using a small number of parameters and less training costs to improve the quality of handheld ultrasound devices.
基金the Postdoctoral ScienceFoundation of China(No.2023M730156)the NationalNatural Foundation of China(No.62301012).
文摘Hyper-and multi-spectral image fusion is an important technology to produce hyper-spectral and hyper-resolution images,which always depends on the spectral response function andthe point spread function.However,few works have been payed on the estimation of the two degra-dation functions.To learn the two functions from image pairs to be fused,we propose a Dirichletnetwork,where both functions are properly constrained.Specifically,the spatial response function isconstrained with positivity,while the Dirichlet distribution along with a total variation is imposedon the point spread function.To the best of our knowledge,the neural network and the Dirichlet regularization are exclusively investigated,for the first time,to estimate the degradation functions.Both image degradation and fusion experiments demonstrate the effectiveness and superiority of theproposed Dirichlet network.
基金the National Natural Science Foundation of China(No.61975015)the Research and Innovation Project for Graduate Students at Zhongyuan University of Technology(No.YKY2024ZK14).
文摘Human posture estimation is a prominent research topic in the fields of human-com-puter interaction,motion recognition,and other intelligent applications.However,achieving highaccuracy in key point localization,which is crucial for intelligent applications,contradicts the lowdetection accuracy of human posture detection models in practical scenarios.To address this issue,a human pose estimation network called AT-HRNet has been proposed,which combines convolu-tional self-attention and cross-dimensional feature transformation.AT-HRNet captures significantfeature information from various regions in an adaptive manner,aggregating them through convolu-tional operations within the local receptive domain.The residual structures TripNeck and Trip-Block of the high-resolution network are designed to further refine the key point locations,wherethe attention weight is adjusted by a cross-dimensional interaction to obtain more features.To vali-date the effectiveness of this network,AT-HRNet was evaluated using the COCO2017 dataset.Theresults show that AT-HRNet outperforms HRNet by improving 3.2%in mAP,4.0%in AP75,and3.9%in AP^(M).This suggests that AT-HRNet can offer more beneficial solutions for human posture estimation.
