This article presents a digital demodulation technique for use in galvanic-coupling intrabody communication (GC IBC) with binary phase shift keying (BPSK)-modulated signals at low frequencies (below megahertz). The technique relies on a direct oversampled acquisition of the BPSK-modulated signal with an analog-to-digital converter (ADC) associated with an original digital processing algorithm to extract the demodulated data from the acquired samples. The originality of the solution resides in the digital processing algorithm, which combines several mechanisms to fully exploit the redundancy present in the collected samples in order to provide a high degree of robustness, while maintaining a low level of complexity compatible with efficient implementation in a microcontroller. Simulation and measurement results are presented, confirming the robustness of the proposed solution. In particular, hardware measurements carried out under controlled conditions demonstrate very good performance, with a bit error rate (BER) below 3 × 10−5 for a signal with a signal-to-noise ratio (SNR) of −5 dB. The proposed solution is also validated under real conditions with a galvanic-coupling (GC) communication realized through the back muscle of a fish, resulting in a BER < 2.5 × 10−6.
{"title":"Implementation and Performance Analysis of a Digital BPSK Demodulation Technique for Galvanic-Coupling Communication","authors":"Stephane Pitou;Vincent Kerzerho;Serge Bernard;Tristan Rouyer;Fabien Soulier;David McKenzie;Florence Azais","doi":"10.1109/JSEN.2025.3641296","DOIUrl":"https://doi.org/10.1109/JSEN.2025.3641296","url":null,"abstract":"This article presents a digital demodulation technique for use in galvanic-coupling intrabody communication (GC IBC) with binary phase shift keying (BPSK)-modulated signals at low frequencies (below megahertz). The technique relies on a direct oversampled acquisition of the BPSK-modulated signal with an analog-to-digital converter (ADC) associated with an original digital processing algorithm to extract the demodulated data from the acquired samples. The originality of the solution resides in the digital processing algorithm, which combines several mechanisms to fully exploit the redundancy present in the collected samples in order to provide a high degree of robustness, while maintaining a low level of complexity compatible with efficient implementation in a microcontroller. Simulation and measurement results are presented, confirming the robustness of the proposed solution. In particular, hardware measurements carried out under controlled conditions demonstrate very good performance, with a bit error rate (BER) below 3 × 10−5 for a signal with a signal-to-noise ratio (SNR) of −5 dB. The proposed solution is also validated under real conditions with a galvanic-coupling (GC) communication realized through the back muscle of a fish, resulting in a BER < 2.5 × 10−6.","PeriodicalId":447,"journal":{"name":"IEEE Sensors Journal","volume":"26 3","pages":"5141-5150"},"PeriodicalIF":4.3,"publicationDate":"2025-12-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146082059","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-17DOI: 10.1109/JSEN.2025.3642263
Xinghua Liu;Jiaxuan Du;Xiang Gao;Shiping Wen;Badong Chen;Peng Wang
We conducted an in-depth investigation into the impact of conditional variational autoencoders (CVAEs) and Bayesian neural networks (BNNs) on high dynamic range (HDR) image reconstruction. A parallel multiattention module (PMAM) is introduced in the improved YOLOv10 framework to enhance computational efficiency and detection performance. To reconstruct HDR images, we enhance the asynchronous Kalman filter (AKF) algorithm to improve image detail quality. We introduce BNN and CVAE into the AKF algorithm to reduce noise effects and improve the logarithmic intensity of the reconstructed image. The BNN estimates noise covariance, thereby reducing its impact during the reconstruction process. Simultaneously, the CVAE leverages polarity as a conditional input, and uses spatial and temporal information through a CVAE to generate more accurate logarithmic image intensities. In the object detection stage, we integrate a parallel module combining the self-attention mechanism and the ECA module to improve training efficiency without increasing the number of parameters. This PMAM module, based on improved YOLOv10, strengthens the model’s ability to capture global and channel-specific features. Finally, the proposed method’s accuracy and robustness are validated through extensive simulations and comparative experiments. Comprehensive experiments on public datasets show that our model achieves 81.32% mAP@0.5 and 59.67% mAP@[0.5:0.95], demonstrating significant improvements in detection accuracy and image reconstruction quality.
{"title":"Event Camera Object Detection Using Bayesian Neural Network-Conditional Variational Autoencoders and Improved YOLOv10","authors":"Xinghua Liu;Jiaxuan Du;Xiang Gao;Shiping Wen;Badong Chen;Peng Wang","doi":"10.1109/JSEN.2025.3642263","DOIUrl":"https://doi.org/10.1109/JSEN.2025.3642263","url":null,"abstract":"We conducted an in-depth investigation into the impact of conditional variational autoencoders (CVAEs) and Bayesian neural networks (BNNs) on high dynamic range (HDR) image reconstruction. A parallel multiattention module (PMAM) is introduced in the improved YOLOv10 framework to enhance computational efficiency and detection performance. To reconstruct HDR images, we enhance the asynchronous Kalman filter (AKF) algorithm to improve image detail quality. We introduce BNN and CVAE into the AKF algorithm to reduce noise effects and improve the logarithmic intensity of the reconstructed image. The BNN estimates noise covariance, thereby reducing its impact during the reconstruction process. Simultaneously, the CVAE leverages polarity as a conditional input, and uses spatial and temporal information through a CVAE to generate more accurate logarithmic image intensities. In the object detection stage, we integrate a parallel module combining the self-attention mechanism and the ECA module to improve training efficiency without increasing the number of parameters. This PMAM module, based on improved YOLOv10, strengthens the model’s ability to capture global and channel-specific features. Finally, the proposed method’s accuracy and robustness are validated through extensive simulations and comparative experiments. Comprehensive experiments on public datasets show that our model achieves 81.32% mAP@0.5 and 59.67% mAP@[0.5:0.95], demonstrating significant improvements in detection accuracy and image reconstruction quality.","PeriodicalId":447,"journal":{"name":"IEEE Sensors Journal","volume":"26 3","pages":"5151-5164"},"PeriodicalIF":4.3,"publicationDate":"2025-12-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146082035","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-12DOI: 10.1109/JSEN.2025.3640716
Xiaoman Guo;Zishi Shen;Xianmin Chen;Xinkun Zhou;Gang Chen;Hu Sun
To address the challenge of cross-region feature distribution shifts in corrosion damage monitoring using ultrasonic-guided wave, this study proposes a transfer learning method based on convolutional neural networks and representation subspace distance (CNNs-RSD). This method aims to improve damage localization accuracy and generalization capability of guided wave signals across diverse structures. This approach extracts damage-sensitive features from Lamb wave in various structures, providing a foundation for subsequent domain adaptation regression (DAR). Simultaneously, representation subspace distance (RSD) is introduced as the domain-adaptive regression module to minimize the geometric distance between source and target feature subspaces from a subspace alignment perspective, which effectively mitigating the regression performance degradation caused by scale perturbation in traditional feature alignment method. To validate the effectiveness of the proposed method, a corrosion damage dataset based on aluminum plate was constructed, and multiple transfer experiments were designed. Damage data corresponding to a 20 mm defect from aluminum plate sample 1 were used as the source domain, and cross-domain recognition tests were subsequently conducted on aluminum plate sample 2 with four different damage sizes (15, 20, 25, and 30 mm). Furthermore, additional validation was performed on two new aluminum plates containing real corrosion defects. The results demonstrate that the CNN-RSD method outperforms comparative models, including 1-D CNN (1D-CNN), CNN-KGW, and gMLP, in terms of mean absolute error (MAE) and localization relative error (LRE), exhibiting superior positioning accuracy and robustness. It also maintains robust positioning performance in real-damage verification, thereby highlighting its cross-domain transferability and potential for engineering applications.
{"title":"A Cross-Domain Corrosion Identification Algorithm of Aircraft Structures Based on Domain Adaptation Regression","authors":"Xiaoman Guo;Zishi Shen;Xianmin Chen;Xinkun Zhou;Gang Chen;Hu Sun","doi":"10.1109/JSEN.2025.3640716","DOIUrl":"https://doi.org/10.1109/JSEN.2025.3640716","url":null,"abstract":"To address the challenge of cross-region feature distribution shifts in corrosion damage monitoring using ultrasonic-guided wave, this study proposes a transfer learning method based on convolutional neural networks and representation subspace distance (CNNs-RSD). This method aims to improve damage localization accuracy and generalization capability of guided wave signals across diverse structures. This approach extracts damage-sensitive features from Lamb wave in various structures, providing a foundation for subsequent domain adaptation regression (DAR). Simultaneously, representation subspace distance (RSD) is introduced as the domain-adaptive regression module to minimize the geometric distance between source and target feature subspaces from a subspace alignment perspective, which effectively mitigating the regression performance degradation caused by scale perturbation in traditional feature alignment method. To validate the effectiveness of the proposed method, a corrosion damage dataset based on aluminum plate was constructed, and multiple transfer experiments were designed. Damage data corresponding to a 20 mm defect from aluminum plate sample 1 were used as the source domain, and cross-domain recognition tests were subsequently conducted on aluminum plate sample 2 with four different damage sizes (15, 20, 25, and 30 mm). Furthermore, additional validation was performed on two new aluminum plates containing real corrosion defects. The results demonstrate that the CNN-RSD method outperforms comparative models, including 1-D CNN (1D-CNN), CNN-KGW, and gMLP, in terms of mean absolute error (MAE) and localization relative error (LRE), exhibiting superior positioning accuracy and robustness. It also maintains robust positioning performance in real-damage verification, thereby highlighting its cross-domain transferability and potential for engineering applications.","PeriodicalId":447,"journal":{"name":"IEEE Sensors Journal","volume":"26 2","pages":"3389-3399"},"PeriodicalIF":4.3,"publicationDate":"2025-12-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145982244","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-11DOI: 10.1109/JSEN.2025.3640827
Kenesbaeva Periyzat Ismaylovna;Azimbek Khudoyberdiev;Hee-Cheol Kim
Traditional mental workload (MW) classification methods often rely on handcrafted features and achieve modest accuracy (70%–85%) while focusing on single modalities or static fusion, thus missing complementary information across sensors. Recent multimodal fusion approaches, such as attention-based weighting, averaging, or majority voting, often fail to accurately assess the relative informativeness of each modality, especially when one sensor becomes unreliable. We introduce CogniMoE, an end-to-end multimodal framework that learns from raw physiological signals with three innovations: 1) a high-efficiency on-the-fly scalogram generation pipeline using FP16 arithmetic that overcomes traditional storage bottlenecks reducing disk space usage by 98% while enabling seamless GPU processing; 2) parallel per-modality CNN–LSTM branches with attention and dynamic dropout that robustly extract modality-specific spatial–temporal features, outperforming single-stream encoders; and 3) an interpretable mixture of experts (MoE) gating mechanism that replaces static fusion with instance-level adaptive weighting, ensuring robustness by dynamically suppressing unreliable modalities in real time. Evaluations on the MAUS, CLAS, and WESAD datasets demonstrate that CogniMoE consistently outperforms both traditional methods (with average accuracies of 70%–85%) and recent state-ofthe- art (SOTA) approaches (up to 92% accuracy), achieving accuracies of 94%, 92%, and 98%, respectively. In addition, the MoE gating mechanism improves classification accuracy by approximately 5% on average over nonadaptive fusion strategies while dynamically adjusting modality importance based on individual participant characteristics.
{"title":"CogniMoE: End-to-End Multimodal Mental Workload Classification via On-the-Fly Scalogram Generation and MoE Gating","authors":"Kenesbaeva Periyzat Ismaylovna;Azimbek Khudoyberdiev;Hee-Cheol Kim","doi":"10.1109/JSEN.2025.3640827","DOIUrl":"https://doi.org/10.1109/JSEN.2025.3640827","url":null,"abstract":"Traditional mental workload (MW) classification methods often rely on handcrafted features and achieve modest accuracy (70%–85%) while focusing on single modalities or static fusion, thus missing complementary information across sensors. Recent multimodal fusion approaches, such as attention-based weighting, averaging, or majority voting, often fail to accurately assess the relative informativeness of each modality, especially when one sensor becomes unreliable. We introduce CogniMoE, an end-to-end multimodal framework that learns from raw physiological signals with three innovations: 1) a high-efficiency on-the-fly scalogram generation pipeline using FP16 arithmetic that overcomes traditional storage bottlenecks reducing disk space usage by 98% while enabling seamless GPU processing; 2) parallel per-modality CNN–LSTM branches with attention and dynamic dropout that robustly extract modality-specific spatial–temporal features, outperforming single-stream encoders; and 3) an interpretable mixture of experts (MoE) gating mechanism that replaces static fusion with instance-level adaptive weighting, ensuring robustness by dynamically suppressing unreliable modalities in real time. Evaluations on the MAUS, CLAS, and WESAD datasets demonstrate that CogniMoE consistently outperforms both traditional methods (with average accuracies of 70%–85%) and recent state-ofthe- art (SOTA) approaches (up to 92% accuracy), achieving accuracies of 94%, 92%, and 98%, respectively. In addition, the MoE gating mechanism improves classification accuracy by approximately 5% on average over nonadaptive fusion strategies while dynamically adjusting modality importance based on individual participant characteristics.","PeriodicalId":447,"journal":{"name":"IEEE Sensors Journal","volume":"26 3","pages":"5213-5228"},"PeriodicalIF":4.3,"publicationDate":"2025-12-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11298422","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146082139","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-11-18DOI: 10.1109/JSEN.2025.3631864
Naga Srinivasarao Chilamkurthy;Shaik Abdul Hakeem;Sreenivasulu Tupakula;Sunil Chinnadurai;Om Jee Pandey;Anirban Ghosh
The rapid expansion of Internet of Things (IoT) applications has driven advancements in networking technologies such as low-power wide-area networks (LPWANs) to extend coverage and enhance the lifespan of IoT devices (IoDs). However, real-world IoT networks are typically heterogeneous, comprising static and dynamic IoDs leading to variations in network topology. These fluctuations cause challenges such as increased data latency and energy imbalances, which hinder efficient information flow. To overcome these issues, this article presents a novel approach that integrates small-world characteristics (SWCs), inspired by social network theory, into heterogeneous LPWANs using reinforcement learning (RL). Specifically, the Q-learning technique is used to introduce new long-range links into the network, enhancing connectivity and optimizing performance. Different conventional networks with varying numbers of mobile nodes are studied in this work followed by their subsequent transformation to small-world versions. The performance of the networks is optimized in terms of energy efficiency and latency in data routing. It is observed that irrespective of the network (conventional or small-world), the performance is better if the number of static nodes is greater. Furthermore, independent of the degree of dynamicity, the SW-LPWAN is more energy-efficient and has lower transmission delay than the corresponding conventional network. Numerically, SWLPWANs achieve up to 14.6% faster data transmission speeds, supporting 19.7% more active IoDs, and maintaining 15.5% higher residual energy compared with conventional networks.
{"title":"Improving the Performance of Heterogeneous LPWANs: An Integrated Small-World and Machine Learning Approach","authors":"Naga Srinivasarao Chilamkurthy;Shaik Abdul Hakeem;Sreenivasulu Tupakula;Sunil Chinnadurai;Om Jee Pandey;Anirban Ghosh","doi":"10.1109/JSEN.2025.3631864","DOIUrl":"https://doi.org/10.1109/JSEN.2025.3631864","url":null,"abstract":"The rapid expansion of Internet of Things (IoT) applications has driven advancements in networking technologies such as low-power wide-area networks (LPWANs) to extend coverage and enhance the lifespan of IoT devices (IoDs). However, real-world IoT networks are typically heterogeneous, comprising static and dynamic IoDs leading to variations in network topology. These fluctuations cause challenges such as increased data latency and energy imbalances, which hinder efficient information flow. To overcome these issues, this article presents a novel approach that integrates small-world characteristics (SWCs), inspired by social network theory, into heterogeneous LPWANs using reinforcement learning (RL). Specifically, the Q-learning technique is used to introduce new long-range links into the network, enhancing connectivity and optimizing performance. Different conventional networks with varying numbers of mobile nodes are studied in this work followed by their subsequent transformation to small-world versions. The performance of the networks is optimized in terms of energy efficiency and latency in data routing. It is observed that irrespective of the network (conventional or small-world), the performance is better if the number of static nodes is greater. Furthermore, independent of the degree of dynamicity, the SW-LPWAN is more energy-efficient and has lower transmission delay than the corresponding conventional network. Numerically, SWLPWANs achieve up to 14.6% faster data transmission speeds, supporting 19.7% more active IoDs, and maintaining 15.5% higher residual energy compared with conventional networks.","PeriodicalId":447,"journal":{"name":"IEEE Sensors Journal","volume":"26 1","pages":"1410-1419"},"PeriodicalIF":4.3,"publicationDate":"2025-11-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145852503","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-11-17DOI: 10.1109/JSEN.2025.3631094
Sihan Chen;Hui Chen;Shuqi Liu;Yige Zhao;Wanquan Liu
Outdoor depth acquisition with technologies like Light Detection and Ranging (LiDAR) is a challenging task due to factors such as high complexity and sensitivity to light variation, which result in sparse point cloud density. This research is an attempt to address these issues and suggests the use of Red, Green, Blue (RGB) image guidance for depth completion of sparse laser point clouds. The presented research work involves three stages. First, to overcome incomplete or inaccurate completion caused by unclear information corresponding RGB image and depth image, a guided convolutional module and a two-stage attention mechanism based on a feature fusion strategy are proposed. The strategy uses lightweight network models to improve the completion accuracy. Second, a completion method based on the fine-grained convolutional space propagation network is proposed to preserve the original depth value and refine the depth map. This scheme addresses the issue of losing the original depth value due to the noise while fusing two different information input modes of RGB image and depth map. Finally, in order to test the depth completion performance of TFDCNet, evaluation is performed by using the KITTI dataset. Experimental results reveal that TFDCNet shows improved completion accuracy by 8.36% in the selected scenarios compared with the state-of-the-art.
{"title":"TFDCNet: Two-Stage Multimodal Fusion and Fine-Grained Convolutional Space Propagation Network for Depth Completion of Outdoor Scenes","authors":"Sihan Chen;Hui Chen;Shuqi Liu;Yige Zhao;Wanquan Liu","doi":"10.1109/JSEN.2025.3631094","DOIUrl":"https://doi.org/10.1109/JSEN.2025.3631094","url":null,"abstract":"Outdoor depth acquisition with technologies like Light Detection and Ranging (LiDAR) is a challenging task due to factors such as high complexity and sensitivity to light variation, which result in sparse point cloud density. This research is an attempt to address these issues and suggests the use of Red, Green, Blue (RGB) image guidance for depth completion of sparse laser point clouds. The presented research work involves three stages. First, to overcome incomplete or inaccurate completion caused by unclear information corresponding RGB image and depth image, a guided convolutional module and a two-stage attention mechanism based on a feature fusion strategy are proposed. The strategy uses lightweight network models to improve the completion accuracy. Second, a completion method based on the fine-grained convolutional space propagation network is proposed to preserve the original depth value and refine the depth map. This scheme addresses the issue of losing the original depth value due to the noise while fusing two different information input modes of RGB image and depth map. Finally, in order to test the depth completion performance of TFDCNet, evaluation is performed by using the KITTI dataset. Experimental results reveal that TFDCNet shows improved completion accuracy by 8.36% in the selected scenarios compared with the state-of-the-art.","PeriodicalId":447,"journal":{"name":"IEEE Sensors Journal","volume":"26 1","pages":"1395-1409"},"PeriodicalIF":4.3,"publicationDate":"2025-11-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145852467","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：