Pub Date : 2024-04-02DOI: 10.1109/TETCI.2024.3378331
Hailin Zou;Zijie Chen;Jing Zhang;Lei Wang;Fuchun Zhang;Jianqing Li;Yuanyuan Pan
Using wearable sensors to identify human activities has elicited significant interest within the discipline of ubiquitous computing for everyday facilitation. Recent research has employed hybrid models to better leverage the modal information of sensors and temporal information, enabling improved performance for wearable human activity recognition. Nevertheless, the lack of effective exploitation of human structural information and limited capacity for cross-channel fusion remains a major challenge. This study proposes a generic design, called GT-WHAR, to accommodate the varying application scenarios and datasets while performing effective feature extraction and fusion. Firstly, a novel and unified representation paradigm, namely Body-Sensing Graph Representation , has been proposed to represent body movement by a graph set, which incorporates structural information by considering the intrinsic connectivity of the skeletal structure. Secondly, the newly designed Body-Node Attention Graph Network employs graph neural networks to extract and fuse the cross-channel information within the graph set. Eventually, the graph network has been embedded in the proposed Bidirectional Temporal Learning Network , facilitating the extraction of temporal information in conjunction with the learned structural features. GT-WHAR outperformed the state-of-the-art methods in extensive experiments conducted on benchmark datasets, proving its validity and efficacy. Besides, we have demonstrated the generality of the framework through multiple research questions and provided an in-depth investigation of various influential factors.
{"title":"GT-WHAR: A Generic Graph-Based Temporal Framework for Wearable Human Activity Recognition With Multiple Sensors","authors":"Hailin Zou;Zijie Chen;Jing Zhang;Lei Wang;Fuchun Zhang;Jianqing Li;Yuanyuan Pan","doi":"10.1109/TETCI.2024.3378331","DOIUrl":"https://doi.org/10.1109/TETCI.2024.3378331","url":null,"abstract":"Using wearable sensors to identify human activities has elicited significant interest within the discipline of ubiquitous computing for everyday facilitation. Recent research has employed hybrid models to better leverage the modal information of sensors and temporal information, enabling improved performance for wearable human activity recognition. Nevertheless, the lack of effective exploitation of human structural information and limited capacity for cross-channel fusion remains a major challenge. This study proposes a generic design, called GT-WHAR, to accommodate the varying application scenarios and datasets while performing effective feature extraction and fusion. Firstly, a novel and unified representation paradigm, namely \u0000<italic>Body-Sensing Graph Representation</i>\u0000, has been proposed to represent body movement by a graph set, which incorporates structural information by considering the intrinsic connectivity of the skeletal structure. Secondly, the newly designed \u0000<italic>Body-Node Attention Graph Network</i>\u0000 employs graph neural networks to extract and fuse the cross-channel information within the graph set. Eventually, the graph network has been embedded in the proposed \u0000<italic>Bidirectional Temporal Learning Network</i>\u0000, facilitating the extraction of temporal information in conjunction with the learned structural features. GT-WHAR outperformed the state-of-the-art methods in extensive experiments conducted on benchmark datasets, proving its validity and efficacy. Besides, we have demonstrated the generality of the framework through multiple research questions and provided an in-depth investigation of various influential factors.","PeriodicalId":13135,"journal":{"name":"IEEE Transactions on Emerging Topics in Computational Intelligence","volume":"8 6","pages":"3912-3924"},"PeriodicalIF":5.3,"publicationDate":"2024-04-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142691753","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-04-02DOI: 10.1109/TETCI.2024.3377671
Mohsen Saffari;Mahdi Khodayar;Mohammad E. Khodayar
Due to the increasing demand for electricity and the inherent uncertainty in power generation, finding efficient solutions to the stochastic alternating current optimal power flow (AC-OPF) problem has become crucial. However, the nonlinear and non-convex nature of AC-OPF, coupled with the growing stochasticity resulting from the integration of renewable energy sources, presents significant challenges in achieving fast and reliable solutions. To address these challenges, this study proposes a novel graph-based generative methodology that effectively captures the uncertainties in power system measurements, enabling the learning of probability distribution functions for generation dispatch and voltage setpoints. Our approach involves modeling the power system as a weighted graph and utilizing a deep spectral graph convolution network to extract powerful spatial patterns from the input graph measurements. A unique variational approach is introduced to identify the most relevant latent features that accurately describe the setpoints of the AC-OPF problem. Additionally, a capsule network with a new greedy dynamic routing algorithm is proposed to precisely decode the latent features and estimate the probabilistic AC-OPF problem. Further, a set of carefully designed physics-informed loss functions is incorporated in the training procedure of the model to ensure adherence to the fundamental physics rules governing power systems. Notably, the proposed physics-informed loss functions not only enhance the accuracy of AC-OPF estimation by effectively regularizing the deep learning model but also significantly reduce the time complexity. Extensive experimental evaluations conducted on various benchmarks demonstrate our proposed model's superiority over both probabilistic and deterministic approaches in terms of relevant criteria.
{"title":"Physics-Informed Graph Capsule Generative Autoencoder for Probabilistic AC Optimal Power Flow","authors":"Mohsen Saffari;Mahdi Khodayar;Mohammad E. Khodayar","doi":"10.1109/TETCI.2024.3377671","DOIUrl":"https://doi.org/10.1109/TETCI.2024.3377671","url":null,"abstract":"Due to the increasing demand for electricity and the inherent uncertainty in power generation, finding efficient solutions to the stochastic alternating current optimal power flow (AC-OPF) problem has become crucial. However, the nonlinear and non-convex nature of AC-OPF, coupled with the growing stochasticity resulting from the integration of renewable energy sources, presents significant challenges in achieving fast and reliable solutions. To address these challenges, this study proposes a novel graph-based generative methodology that effectively captures the uncertainties in power system measurements, enabling the learning of probability distribution functions for generation dispatch and voltage setpoints. Our approach involves modeling the power system as a weighted graph and utilizing a deep spectral graph convolution network to extract powerful spatial patterns from the input graph measurements. A unique variational approach is introduced to identify the most relevant latent features that accurately describe the setpoints of the AC-OPF problem. Additionally, a capsule network with a new greedy dynamic routing algorithm is proposed to precisely decode the latent features and estimate the probabilistic AC-OPF problem. Further, a set of carefully designed physics-informed loss functions is incorporated in the training procedure of the model to ensure adherence to the fundamental physics rules governing power systems. Notably, the proposed physics-informed loss functions not only enhance the accuracy of AC-OPF estimation by effectively regularizing the deep learning model but also significantly reduce the time complexity. Extensive experimental evaluations conducted on various benchmarks demonstrate our proposed model's superiority over both probabilistic and deterministic approaches in terms of relevant criteria.","PeriodicalId":13135,"journal":{"name":"IEEE Transactions on Emerging Topics in Computational Intelligence","volume":"8 5","pages":"3382-3395"},"PeriodicalIF":5.3,"publicationDate":"2024-04-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142368573","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-04-02DOI: 10.1109/TETCI.2024.3367811
Jie Yang;Chin-Teng Lin
Hierarchical clustering is able to provide partitions of different granularity levels. However, most existing hierarchical clustering techniques perform clustering in the original feature space of the data, which may suffer from overlap, sparseness, or other undesirable characteristics, resulting in noncompetitive performance. In the field of deep clustering, learning representations using pseudo labels has recently become a research hotspot. Yet most existing approaches employ coarse-grained pseudo labels, which may contain noise or incorrect labels. Hence, the learned feature space does not produce a competitive model. In this paper, we introduce the idea of fine-grained labels of supervised learning into unsupervised clustering, giving rise to the enhanced adjacency-constrained hierarchical clustering (ECHC) model. The full framework comprises four steps. One, adjacency-constrained hierarchical clustering (CHC) is used to produce relatively pure fine-grained pseudo labels. Two, those fine-grained pseudo labels are used to train a shallow multilayer perceptron to generate good representations. Three, the corresponding representation of each sample in the learned space is used to construct a similarity matrix. Four, CHC is used to generate the final partition based on the similarity matrix. The experimental results show that the proposed ECHC framework not only outperforms 14 shallow clustering methods on eight real-world datasets but also surpasses current state-of-the-art deep clustering models on six real-world datasets. In addition, on five real-world datasets, ECHC achieves comparable results to supervised algorithms.
{"title":"Enhanced Adjacency-Constrained Hierarchical Clustering Using Fine-Grained Pseudo Labels","authors":"Jie Yang;Chin-Teng Lin","doi":"10.1109/TETCI.2024.3367811","DOIUrl":"https://doi.org/10.1109/TETCI.2024.3367811","url":null,"abstract":"Hierarchical clustering is able to provide partitions of different granularity levels. However, most existing hierarchical clustering techniques perform clustering in the original feature space of the data, which may suffer from overlap, sparseness, or other undesirable characteristics, resulting in noncompetitive performance. In the field of deep clustering, learning representations using pseudo labels has recently become a research hotspot. Yet most existing approaches employ coarse-grained pseudo labels, which may contain noise or incorrect labels. Hence, the learned feature space does not produce a competitive model. In this paper, we introduce the idea of fine-grained labels of supervised learning into unsupervised clustering, giving rise to the enhanced adjacency-constrained hierarchical clustering (ECHC) model. The full framework comprises four steps. One, adjacency-constrained hierarchical clustering (CHC) is used to produce relatively pure fine-grained pseudo labels. Two, those fine-grained pseudo labels are used to train a shallow multilayer perceptron to generate good representations. Three, the corresponding representation of each sample in the learned space is used to construct a similarity matrix. Four, CHC is used to generate the final partition based on the similarity matrix. The experimental results show that the proposed ECHC framework not only outperforms 14 shallow clustering methods on eight real-world datasets but also surpasses current state-of-the-art deep clustering models on six real-world datasets. In addition, on five real-world datasets, ECHC achieves comparable results to supervised algorithms.","PeriodicalId":13135,"journal":{"name":"IEEE Transactions on Emerging Topics in Computational Intelligence","volume":"8 3","pages":"2481-2492"},"PeriodicalIF":5.3,"publicationDate":"2024-04-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141096325","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-04-02DOI: 10.1109/TETCI.2024.3377728
Jingke Yan;Yao Cheng;Qin Wang;Lei Liu;Weihua Zhang;Bo Jin
Thanks to the development of deep learning, machine abnormal sound detection (MASD) based on unsupervised learning has exhibited excellent performance. However, in the task of unsupervised MASD, there are discrepancies between the acoustic characteristics of the test set and the training set under the physical parameter changes (domain shifts) of the same machine's operating conditions. Existing methods not only struggle to stably learn the sound signal features under various domain shifts but also inevitably increase computational overhead. To address these issues, we propose an unsupervised machine abnormal sound detection model based on Transformer and Dynamic Graph Convolution (Unsuper-TDGCN) in this paper. Firstly, we design a network that models time-frequency domain features to capture both global and local spatial and time-frequency interactions, thus improving the model's stability under domain shifts. Then, we introduce a Dynamic Graph Convolutional Network (DyGCN) to model the dependencies between features under domain shifts, enhancing the model's ability to perceive changes in domain features. Finally, a Domain Self-adaptive Network (DSN) is employed to compensate for the performance decline caused by domain shifts, thereby improving the model's adaptive ability for detecting anomalous sounds in MASD tasks under domain shifts. The effectiveness of our proposed model has been validated on multiple datasets.
{"title":"Transformer and Graph Convolution-Based Unsupervised Detection of Machine Anomalous Sound Under Domain Shifts","authors":"Jingke Yan;Yao Cheng;Qin Wang;Lei Liu;Weihua Zhang;Bo Jin","doi":"10.1109/TETCI.2024.3377728","DOIUrl":"https://doi.org/10.1109/TETCI.2024.3377728","url":null,"abstract":"Thanks to the development of deep learning, machine abnormal sound detection (MASD) based on unsupervised learning has exhibited excellent performance. However, in the task of unsupervised MASD, there are discrepancies between the acoustic characteristics of the test set and the training set under the physical parameter changes (domain shifts) of the same machine's operating conditions. Existing methods not only struggle to stably learn the sound signal features under various domain shifts but also inevitably increase computational overhead. To address these issues, we propose an unsupervised machine abnormal sound detection model based on Transformer and Dynamic Graph Convolution (Unsuper-TDGCN) in this paper. Firstly, we design a network that models time-frequency domain features to capture both global and local spatial and time-frequency interactions, thus improving the model's stability under domain shifts. Then, we introduce a Dynamic Graph Convolutional Network (DyGCN) to model the dependencies between features under domain shifts, enhancing the model's ability to perceive changes in domain features. Finally, a Domain Self-adaptive Network (DSN) is employed to compensate for the performance decline caused by domain shifts, thereby improving the model's adaptive ability for detecting anomalous sounds in MASD tasks under domain shifts. The effectiveness of our proposed model has been validated on multiple datasets.","PeriodicalId":13135,"journal":{"name":"IEEE Transactions on Emerging Topics in Computational Intelligence","volume":"8 4","pages":"2827-2842"},"PeriodicalIF":5.3,"publicationDate":"2024-04-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141965845","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-04-02DOI: 10.1109/TETCI.2024.3378649
Lianyu Hu;Liqing Gao;Zekang Liu;Wei Feng
Despite the recent progress of continuous sign language recognition (CSLR), most state-of-the-art methods process input sign language videos frame by frame to predict sentences. This usually causes a heavy computational burden and is inefficient and even infeasible in real-world scenarios. Inspired by the fact that videos are inherently redundant where not all frames are essential for recognition, we propose spatial temporal aggregation (STAgg) to address this problem. Specifically, STAgg synthesizes adjacent similar frames into a unified robust representation before being fed into the recognition module, thus highly reducing the computation complexity and memory demand. We first give a detailed analysis on commonly-used aggregation methods like subsampling, max pooling and average, and then naturally derive our STAgg from the expected design criterion. Compared to commonly used pooling and subsampling counterparts, extensive ablation studies verify the superiority of our proposed three diverse STAgg variants in both accuracy and efficiency. The best version achieves comparative accuracy with state-of-the-art competitors, but is 1.35× faster with only 0.50× computational costs, consuming 0.70× training time and 0.65× memory usage. Experiments on four large-scale datasets upon multiple backbones fully verify the generalizability and effectiveness of the proposed STAgg. Another advantage of STAgg is enabling more powerful backbones, which may further boost the accuracy of CSLR under similar computational/memory budgets. We also visualize the results of STAgg to support intuitive and insightful analysis of the effects of STAgg.
{"title":"Spatial Temporal Aggregation for Efficient Continuous Sign Language Recognition","authors":"Lianyu Hu;Liqing Gao;Zekang Liu;Wei Feng","doi":"10.1109/TETCI.2024.3378649","DOIUrl":"https://doi.org/10.1109/TETCI.2024.3378649","url":null,"abstract":"Despite the recent progress of continuous sign language recognition (CSLR), most state-of-the-art methods process input sign language videos frame by frame to predict sentences. This usually causes a heavy computational burden and is inefficient and even infeasible in real-world scenarios. Inspired by the fact that videos are inherently redundant where not all frames are essential for recognition, we propose spatial temporal aggregation (STAgg) to address this problem. Specifically, STAgg synthesizes adjacent similar frames into a unified robust representation before being fed into the recognition module, thus highly reducing the computation complexity and memory demand. We first give a detailed analysis on commonly-used aggregation methods like subsampling, max pooling and average, and then naturally derive our STAgg from the expected design criterion. Compared to commonly used pooling and subsampling counterparts, extensive ablation studies verify the superiority of our proposed three diverse STAgg variants in both accuracy and efficiency. The best version achieves comparative accuracy with state-of-the-art competitors, but is 1.35× faster with only 0.50× computational costs, consuming 0.70× training time and 0.65× memory usage. Experiments on four large-scale datasets upon multiple backbones fully verify the generalizability and effectiveness of the proposed STAgg. Another advantage of STAgg is enabling more powerful backbones, which may further boost the accuracy of CSLR under similar computational/memory budgets. We also visualize the results of STAgg to support intuitive and insightful analysis of the effects of STAgg.","PeriodicalId":13135,"journal":{"name":"IEEE Transactions on Emerging Topics in Computational Intelligence","volume":"8 6","pages":"3925-3935"},"PeriodicalIF":5.3,"publicationDate":"2024-04-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142691765","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-04-01DOI: 10.1109/TETCI.2024.3375450
G. Narayanan;M. Syed Ali;Rajagopal Karthikeyan;Grienggrai Rajchakit;Sumaya Sanober;Pankaj Kumar
This paper addresses the Mittag-Leffler synchronization problem of fractional-order reaction-diffusion complex-valued neural networks (FRDCVNNs) with delays. New Mittag-Leffler synchronization (MLS) criteria in the form of the $p$ -norm for an error model derived from the drive-response model are constructed. In the design of the adaptive feedback controller, the Lyapunov approach is considered in the framework of the $p$ -norm technique, and less conservative algebraic conditions that guarantee MLS for the considered model are given. Moreover, the MLS of the considered model without reaction diffusion effect is investigated using adaptive control. Finally, an example is used to validate the proposed control scheme. To demonstrate the advantages and superiority of the proposed technique over existing methods, an image encryption method based on MLS of FRDCVNNs is considered and solved using the proposed method.
{"title":"Adaptive Strategies and its Application in the Mittag-Leffler Synchronization of Delayed Fractional-Order Complex-Valued Reaction-Diffusion Neural Networks","authors":"G. Narayanan;M. Syed Ali;Rajagopal Karthikeyan;Grienggrai Rajchakit;Sumaya Sanober;Pankaj Kumar","doi":"10.1109/TETCI.2024.3375450","DOIUrl":"https://doi.org/10.1109/TETCI.2024.3375450","url":null,"abstract":"This paper addresses the Mittag-Leffler synchronization problem of fractional-order reaction-diffusion complex-valued neural networks (FRDCVNNs) with delays. New Mittag-Leffler synchronization (MLS) criteria in the form of the \u0000<inline-formula><tex-math>$p$</tex-math></inline-formula>\u0000-norm for an error model derived from the drive-response model are constructed. In the design of the adaptive feedback controller, the Lyapunov approach is considered in the framework of the \u0000<inline-formula><tex-math>$p$</tex-math></inline-formula>\u0000-norm technique, and less conservative algebraic conditions that guarantee MLS for the considered model are given. Moreover, the MLS of the considered model without reaction diffusion effect is investigated using adaptive control. Finally, an example is used to validate the proposed control scheme. To demonstrate the advantages and superiority of the proposed technique over existing methods, an image encryption method based on MLS of FRDCVNNs is considered and solved using the proposed method.","PeriodicalId":13135,"journal":{"name":"IEEE Transactions on Emerging Topics in Computational Intelligence","volume":"8 5","pages":"3294-3307"},"PeriodicalIF":5.3,"publicationDate":"2024-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142368451","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Due to the excellent time-varying problem-solving capability of zeroing neural network (ZNN), many redundancy resolution schemes based on ZNN have been proposed for robots. The work proposes a fixed-time robust ZNN (FTRZNN) model with adaptive parameters to effectively address redundancy resolution problems of robots in the presence of noises. Differing from existing ZNN models, the FTRZNN possesses a fixed-time activation function and two adaptive parameters, which greatly improve its performance on convergence speed and robustness. The establishment of the FTRZNN for handling redundancy resolution problems consists of two steps: 1) converting the target practical problem into nonlinear equations firstly; and 2) deriving an FTRZNN for solving the equations. For providing a convincible evidence of the significant advantages of the FTRZNN over existing ZNN models, theoretical analysis in convergence and robustness of the FTRZNN is given, and the performance of the FTRZNN model is compared with existing ZNN models when performing path tracking tasks using a 6R manipulator under different noise disturbances. Finally, the FTRZNN model is employed to control two robot manipulators (UR5 and Jaco) to track desired paths under noise interference, which is simulated on a robotic simulation platform (i.e.,CoppeliaSim). Simulation results indicate the effectiveness and potential practical value of the FTRZNN model.
{"title":"A Fixed-Time Robust ZNN Model With Adaptive Parameters for Redundancy Resolution of Manipulators","authors":"Mengrui Cao;Lin Xiao;Qiuyue Zuo;Ping Tan;Yongjun He;Xieping Gao","doi":"10.1109/TETCI.2024.3377672","DOIUrl":"https://doi.org/10.1109/TETCI.2024.3377672","url":null,"abstract":"Due to the excellent time-varying problem-solving capability of zeroing neural network (ZNN), many redundancy resolution schemes based on ZNN have been proposed for robots. The work proposes a fixed-time robust ZNN (FTRZNN) model with adaptive parameters to effectively address redundancy resolution problems of robots in the presence of noises. Differing from existing ZNN models, the FTRZNN possesses a fixed-time activation function and two adaptive parameters, which greatly improve its performance on convergence speed and robustness. The establishment of the FTRZNN for handling redundancy resolution problems consists of two steps: 1) converting the target practical problem into nonlinear equations firstly; and 2) deriving an FTRZNN for solving the equations. For providing a convincible evidence of the significant advantages of the FTRZNN over existing ZNN models, theoretical analysis in convergence and robustness of the FTRZNN is given, and the performance of the FTRZNN model is compared with existing ZNN models when performing path tracking tasks using a 6R manipulator under different noise disturbances. Finally, the FTRZNN model is employed to control two robot manipulators (UR5 and Jaco) to track desired paths under noise interference, which is simulated on a robotic simulation platform (i.e.,CoppeliaSim). Simulation results indicate the effectiveness and potential practical value of the FTRZNN model.","PeriodicalId":13135,"journal":{"name":"IEEE Transactions on Emerging Topics in Computational Intelligence","volume":"8 6","pages":"3886-3898"},"PeriodicalIF":5.3,"publicationDate":"2024-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142691732","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-04-01DOI: 10.1109/TETCI.2024.3378599
Ling Huang;Can-Rong Guan;Zhen-Wei Huang;Yuefang Gao;Chang-Dong Wang;C. L. P. Chen
Recently, Deep Neural Networks (DNNs) have been largely utilized in Collaborative Filtering (CF) to produce more accurate recommendation results due to their ability of extracting the nonlinear relationships in the user-item pairs. However, the DNNs-based models usually encounter high computational complexity, i.e., consuming very long training time and storing huge amount of trainable parameters. To address these problems, we develop a novel broad recommender system named Broad Collaborative Filtering (BroadCF), which is an efficient nonlinear collaborative filtering approach. Instead of DNNs, Broad Learning System (BLS) is used as a mapping function to learn the nonlinear matching relationships in the user-item pairs, which can avoid the above issues while achieving very satisfactory rating prediction performance. Contrary to DNNs, BLS is a shallow network that captures nonlinear relationships between input features simply and efficiently. However, directly feeding the original rating data into BLS is not suitable due to the very large dimensionality of the original rating vector. To this end, a new preprocessing procedure is designed to generate user-item rating collaborative vector, which is a low-dimensional user-item input vector that can leverage quality judgments of the most similar users/items. Convincing experimental results on seven datasets have demonstrated the effectiveness of the BroadCF algorithm.
{"title":"Broad Recommender System: An Efficient Nonlinear Collaborative Filtering Approach","authors":"Ling Huang;Can-Rong Guan;Zhen-Wei Huang;Yuefang Gao;Chang-Dong Wang;C. L. P. Chen","doi":"10.1109/TETCI.2024.3378599","DOIUrl":"https://doi.org/10.1109/TETCI.2024.3378599","url":null,"abstract":"Recently, Deep Neural Networks (DNNs) have been largely utilized in Collaborative Filtering (CF) to produce more accurate recommendation results due to their ability of extracting the nonlinear relationships in the user-item pairs. However, the DNNs-based models usually encounter high computational complexity, i.e., consuming very long training time and storing huge amount of trainable parameters. To address these problems, we develop a novel broad recommender system named Broad Collaborative Filtering (BroadCF), which is an efficient nonlinear collaborative filtering approach. Instead of DNNs, Broad Learning System (BLS) is used as a mapping function to learn the nonlinear matching relationships in the user-item pairs, which can avoid the above issues while achieving very satisfactory rating prediction performance. Contrary to DNNs, BLS is a shallow network that captures nonlinear relationships between input features simply and efficiently. However, directly feeding the original rating data into BLS is not suitable due to the very large dimensionality of the original rating vector. To this end, a new preprocessing procedure is designed to generate user-item rating collaborative vector, which is a low-dimensional user-item input vector that can leverage quality judgments of the most similar users/items. Convincing experimental results on seven datasets have demonstrated the effectiveness of the BroadCF algorithm.","PeriodicalId":13135,"journal":{"name":"IEEE Transactions on Emerging Topics in Computational Intelligence","volume":"8 4","pages":"2843-2857"},"PeriodicalIF":5.3,"publicationDate":"2024-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141965846","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-04-01DOI: 10.1109/TETCI.2024.3378651
Yubo Zhao;Jiaqi Wu;Wei Chen;Zehua Wang;Zijian Tian;Fei Richard Yu;Victor C. M. Leung
Power inspection in low-illuminance environments is of great significance for ensuring the all-weather stable operation of the power system. However, low visibility at night seriously interferes with the detection performance of small-sized power devices. In response to the issue, we propose a small object real-time detection method for power line inspection in low-illuminance environments. We design an adaptive transformer-ISP (ATISP) module, in which the optimal parameter regression module generates hyperparameters by sensing input image features to guide the image signal processors (ISPs) to perform image enhancement. With the advantage of ISPs, the ATISP has the advantages of fast inference speed and less training cost. Furthermore, the optimal parameter regression module extracts local features and long-distance dependencies through CNN and Transformer to be able to more fully perceive the input image, so that the generated hyperparameters better enhance image defects. In addition, we use lightweight neural network MobileNetv3 to improve YOLOv7, so that the algorithm maintains excellent small object detection performance while significantly increasing the detection speed. Moreover, the integrated model optimisation uses only the object detection loss functions, which allows ATISP to perform image enhancement just according to the object detection needs, improving small object detection effect and shortening the inference time of ATISP. In extensive experiments, compared with 9 state-of-the-art object detection algorithms, our algorithm has the best small-scale insulator faults detection precision (mAP:75.38 $%$ ) in our DIFE, best small object detection precision (mAP:56.31 $%$ ) in public dataset Exdark, and faster detection speed (FPS:98.81 and 97.53), which prove our method can achieve fast and accurate low-illuminance insulators detection.
{"title":"A Small Object Real-Time Detection Method for Power Line Inspection in Low-Illuminance Environments","authors":"Yubo Zhao;Jiaqi Wu;Wei Chen;Zehua Wang;Zijian Tian;Fei Richard Yu;Victor C. M. Leung","doi":"10.1109/TETCI.2024.3378651","DOIUrl":"https://doi.org/10.1109/TETCI.2024.3378651","url":null,"abstract":"Power inspection in low-illuminance environments is of great significance for ensuring the all-weather stable operation of the power system. However, low visibility at night seriously interferes with the detection performance of small-sized power devices. In response to the issue, we propose a small object real-time detection method for power line inspection in low-illuminance environments. We design an adaptive transformer-ISP (ATISP) module, in which the optimal parameter regression module generates hyperparameters by sensing input image features to guide the image signal processors (ISPs) to perform image enhancement. With the advantage of ISPs, the ATISP has the advantages of fast inference speed and less training cost. Furthermore, the optimal parameter regression module extracts local features and long-distance dependencies through CNN and Transformer to be able to more fully perceive the input image, so that the generated hyperparameters better enhance image defects. In addition, we use lightweight neural network MobileNetv3 to improve YOLOv7, so that the algorithm maintains excellent small object detection performance while significantly increasing the detection speed. Moreover, the integrated model optimisation uses only the object detection loss functions, which allows ATISP to perform image enhancement just according to the object detection needs, improving small object detection effect and shortening the inference time of ATISP. In extensive experiments, compared with 9 state-of-the-art object detection algorithms, our algorithm has the best small-scale insulator faults detection precision (mAP:75.38\u0000<inline-formula><tex-math>$%$</tex-math></inline-formula>\u0000) in our DIFE, best small object detection precision (mAP:56.31\u0000<inline-formula><tex-math>$%$</tex-math></inline-formula>\u0000) in public dataset Exdark, and faster detection speed (FPS:98.81 and 97.53), which prove our method can achieve fast and accurate low-illuminance insulators detection.","PeriodicalId":13135,"journal":{"name":"IEEE Transactions on Emerging Topics in Computational Intelligence","volume":"8 6","pages":"3936-3950"},"PeriodicalIF":5.3,"publicationDate":"2024-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142691755","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-04-01DOI: 10.1109/TETCI.2024.3379165
Yukuan Yang;Qihang Fan;Tianyi Yan;Jing Pei;Guoqi Li
Neuromorphic chips with multi-core architecture are considered to be of great potential for the next generation of artificial intelligence (AI) chips because of the avoidance of the memory wall effect. Deploying deep neural networks (DNNs) to these chips requires two stages, namely, network partition and core placement. For the network partition, existing schemes are mostly manual or only focus on single-layer, small-scale network partitions. For the core placement, to the best of our knowledge, there is still no work that has completely solved the communication deadlock problem at the clock-level which commonly exists in the applications of neuromorphic multi-core and multi-chip (NMCMC) systems. To address these issues that affect the operating and deployment efficiency of NMCMC systems, we formulate the network group partition problem as an optimization problem for the first time and propose a search-based network group partition scheme to solve the problem. A clock-level multi-chip simulator is established to completely avoid the deadlock problem during the core placement optimization process. What's more, a region constrained simulated annealing (RCSA) algorithm is proposed to improve the efficiency of the core placement optimization. Finally, an automated toolchain for the efficient deployment of DNNs in the NMCMC systems is developed by integrating the proposed network group partition and core placement schemes together. Experiments show the proposed group partition scheme can achieve 22.25%, 17.77%, 14.80% less in core number, 9.44%, 7.96%, 5.16% improvements in memory utilization, and more balanced communication and computation loads compared with existing manual schemes in ResNet-18, ResNet-34, and ResNet-50, respectively. In addition, the proposed core placement optimization based on the RCSA algorithm shows higher efficiency with much fewer optimization steps and can realize 9.52%, 11.91%, and 27.52% higher in throughput compared with sequential core placement without deadlock in the ResNet-18, ResNet-34, and ResNet-50 networks. This work paves the way for applying NMCMC systems to real-world scenarios to reach more powerful machine intelligence.
{"title":"Network Group Partition and Core Placement Optimization for Neuromorphic Multi-Core and Multi-Chip Systems","authors":"Yukuan Yang;Qihang Fan;Tianyi Yan;Jing Pei;Guoqi Li","doi":"10.1109/TETCI.2024.3379165","DOIUrl":"https://doi.org/10.1109/TETCI.2024.3379165","url":null,"abstract":"Neuromorphic chips with multi-core architecture are considered to be of great potential for the next generation of artificial intelligence (AI) chips because of the avoidance of the memory wall effect. Deploying deep neural networks (DNNs) to these chips requires two stages, namely, network partition and core placement. For the network partition, existing schemes are mostly manual or only focus on single-layer, small-scale network partitions. For the core placement, to the best of our knowledge, there is still no work that has completely solved the communication deadlock problem at the clock-level which commonly exists in the applications of neuromorphic multi-core and multi-chip (NMCMC) systems. To address these issues that affect the operating and deployment efficiency of NMCMC systems, we formulate the network group partition problem as an optimization problem for the first time and propose a search-based network group partition scheme to solve the problem. A clock-level multi-chip simulator is established to completely avoid the deadlock problem during the core placement optimization process. What's more, a region constrained simulated annealing (RCSA) algorithm is proposed to improve the efficiency of the core placement optimization. Finally, an automated toolchain for the efficient deployment of DNNs in the NMCMC systems is developed by integrating the proposed network group partition and core placement schemes together. Experiments show the proposed group partition scheme can achieve 22.25%, 17.77%, 14.80% less in core number, 9.44%, 7.96%, 5.16% improvements in memory utilization, and more balanced communication and computation loads compared with existing manual schemes in ResNet-18, ResNet-34, and ResNet-50, respectively. In addition, the proposed core placement optimization based on the RCSA algorithm shows higher efficiency with much fewer optimization steps and can realize 9.52%, 11.91%, and 27.52% higher in throughput compared with sequential core placement without deadlock in the ResNet-18, ResNet-34, and ResNet-50 networks. This work paves the way for applying NMCMC systems to real-world scenarios to reach more powerful machine intelligence.","PeriodicalId":13135,"journal":{"name":"IEEE Transactions on Emerging Topics in Computational Intelligence","volume":"8 6","pages":"3966-3981"},"PeriodicalIF":5.3,"publicationDate":"2024-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142691690","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","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：