Pub Date : 2024-04-12DOI: 10.1109/TCDS.2024.3386664
Hao Zhou;Lu Qi;Hai Huang;Xu Yang;Jing Yang
Underwater object detection is challenged by the presence of image blur induced by light absorption and scattering, resulting in substantial performance degradation. It is hypothesized that the attenuation of light is directly correlated with the camera-to-object distance, manifesting as variable degrees of image blur across different regions within underwater images. Specifically, regions in close proximity to the camera exhibit less pronounced blur compared to distant regions. Within the same object category, objects situated in clear regions share similar feature embeddings with their counterparts in blurred regions. This observation underscores the potential for leveraging objects in clear regions to aid in the detection of objects within blurred areas, a critical requirement for autonomous agents, such as autonomous underwater vehicles, engaged in continuous underwater object detection. Motivated by this insight, we introduce the spatiotemporal feature enhancement network (STFEN), a novel framework engineered to autonomously extract discriminative features from objects in clear regions. These features are then harnessed to enhance the representations of objects in blurred regions, operating across both spatial and temporal dimensions. Notably, the proposed STFEN seamlessly integrates into two-stage detectors, such as the faster region-based convolutional neural networks (Faster R-CNN) and feature pyramid networks (FPN). Extensive experimentation conducted on two benchmark underwater datasets, URPC 2018 and URPC 2019, conclusively demonstrates the efficacy of the STFEN framework. It delivers substantial enhancements in performance relative to baseline methods, yielding improvements in the mAP evaluation metric ranging from 3.7% to 5.0%.
{"title":"Spatiotemporal Feature Enhancement Network for Blur Robust Underwater Object Detection","authors":"Hao Zhou;Lu Qi;Hai Huang;Xu Yang;Jing Yang","doi":"10.1109/TCDS.2024.3386664","DOIUrl":"10.1109/TCDS.2024.3386664","url":null,"abstract":"Underwater object detection is challenged by the presence of image blur induced by light absorption and scattering, resulting in substantial performance degradation. It is hypothesized that the attenuation of light is directly correlated with the camera-to-object distance, manifesting as variable degrees of image blur across different regions within underwater images. Specifically, regions in close proximity to the camera exhibit less pronounced blur compared to distant regions. Within the same object category, objects situated in clear regions share similar feature embeddings with their counterparts in blurred regions. This observation underscores the potential for leveraging objects in clear regions to aid in the detection of objects within blurred areas, a critical requirement for autonomous agents, such as autonomous underwater vehicles, engaged in continuous underwater object detection. Motivated by this insight, we introduce the spatiotemporal feature enhancement network (STFEN), a novel framework engineered to autonomously extract discriminative features from objects in clear regions. These features are then harnessed to enhance the representations of objects in blurred regions, operating across both spatial and temporal dimensions. Notably, the proposed STFEN seamlessly integrates into two-stage detectors, such as the faster region-based convolutional neural networks (Faster R-CNN) and feature pyramid networks (FPN). Extensive experimentation conducted on two benchmark underwater datasets, URPC 2018 and URPC 2019, conclusively demonstrates the efficacy of the STFEN framework. It delivers substantial enhancements in performance relative to baseline methods, yielding improvements in the mAP evaluation metric ranging from 3.7% to 5.0%.","PeriodicalId":54300,"journal":{"name":"IEEE Transactions on Cognitive and Developmental Systems","volume":"16 5","pages":"1814-1828"},"PeriodicalIF":5.0,"publicationDate":"2024-04-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140593967","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 : 2024-04-12DOI: 10.1109/tcds.2024.3388152
Qinghui Hong, Qing Li, Jia Li, Jingru Sun, Sichun Du
{"title":"Programmable Bionic Control Circuit Based on Central Pattern Generator","authors":"Qinghui Hong, Qing Li, Jia Li, Jingru Sun, Sichun Du","doi":"10.1109/tcds.2024.3388152","DOIUrl":"https://doi.org/10.1109/tcds.2024.3388152","url":null,"abstract":"","PeriodicalId":54300,"journal":{"name":"IEEE Transactions on Cognitive and Developmental Systems","volume":"64 1","pages":""},"PeriodicalIF":5.0,"publicationDate":"2024-04-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140594171","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 : 2024-04-11DOI: 10.1109/TCDS.2024.3387575
Peng Yu;Ning Tan;Zhaohui Zhong;Cong Hu;Binbin Qiu;Changsheng Li
In modern manufacturing, redundant manipulators have been widely deployed. Performing a task often requires the manipulator to follow specific trajectories while avoiding surrounding obstacles. Different from most existing obstacle-avoidance (OA) schemes that rely on the kinematic model of redundant manipulators, in this article, we propose a new data-driven obstacle-avoidance (DDOA) scheme for the collision-free tracking control of redundant manipulators. The OA task is formulated as a quadratic programming problem with inequality constraints. Then, the objectives of obstacle avoidance and tracking control are unitedly transformed into a computation problem of solving a system including three recurrent neural networks. With the Jacobian estimators designed based on zeroing neural networks, the manipulator Jacobian and critical-point Jacobian can be estimated in a data-driven way without knowing the kinematic model. Finally, the effectiveness of the proposed scheme is validated through extensive simulations and experiments.
{"title":"Unifying Obstacle Avoidance and Tracking Control of Redundant Manipulators Subject to Joint Constraints: A New Data-Driven Scheme","authors":"Peng Yu;Ning Tan;Zhaohui Zhong;Cong Hu;Binbin Qiu;Changsheng Li","doi":"10.1109/TCDS.2024.3387575","DOIUrl":"10.1109/TCDS.2024.3387575","url":null,"abstract":"In modern manufacturing, redundant manipulators have been widely deployed. Performing a task often requires the manipulator to follow specific trajectories while avoiding surrounding obstacles. Different from most existing obstacle-avoidance (OA) schemes that rely on the kinematic model of redundant manipulators, in this article, we propose a new data-driven obstacle-avoidance (DDOA) scheme for the collision-free tracking control of redundant manipulators. The OA task is formulated as a quadratic programming problem with inequality constraints. Then, the objectives of obstacle avoidance and tracking control are unitedly transformed into a computation problem of solving a system including three recurrent neural networks. With the Jacobian estimators designed based on zeroing neural networks, the manipulator Jacobian and critical-point Jacobian can be estimated in a data-driven way without knowing the kinematic model. Finally, the effectiveness of the proposed scheme is validated through extensive simulations and experiments.","PeriodicalId":54300,"journal":{"name":"IEEE Transactions on Cognitive and Developmental Systems","volume":"16 5","pages":"1861-1871"},"PeriodicalIF":5.0,"publicationDate":"2024-04-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140594184","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 : 2024-04-09DOI: 10.1109/TCDS.2024.3386656
Shi Chen;Ming Jiang;Qi Zhao
There is growing interest in understanding the visual behavioral patterns of individuals with autism spectrum disorder (ASD) based on their attentional preferences. Attention reveals the cognitive or perceptual variation in ASD and can serve as a biomarker to assist diagnosis and intervention. The development of machine learning methods for attention-based ASD screening shows promises, yet it has been limited by the need for high-precision eye trackers, the scope of stimuli, and black-box neural networks, making it impractical for real-life clinical scenarios. This study proposes an interpretable and generalizable framework for quantifying atypical attention in people with ASD. Our framework utilizes photos taken by participants with standard cameras to enable practical and flexible deployment in resource-constrained regions. With an emphasis on interpretability and trustworthiness, our method automates human-like diagnostic reasoning, associates photos with semantically plausible attention patterns, and provides clinical evidence to support ASD experts. We further evaluate models on both in-domain and out-of-domain data and demonstrate that our approach accurately classifies individuals with ASD and generalizes across different domains. The proposed method offers an innovative, reliable, and cost-effective tool to assist the diagnostic procedure, which can be an important effort toward transforming clinical research in ASD screening with artificial intelligence systems. Our code is publicly available at https://github.com/szzexpoi/proto_asd .
{"title":"Deep Learning to Interpret Autism Spectrum Disorder Behind the Camera","authors":"Shi Chen;Ming Jiang;Qi Zhao","doi":"10.1109/TCDS.2024.3386656","DOIUrl":"10.1109/TCDS.2024.3386656","url":null,"abstract":"There is growing interest in understanding the visual behavioral patterns of individuals with autism spectrum disorder (ASD) based on their attentional preferences. Attention reveals the cognitive or perceptual variation in ASD and can serve as a biomarker to assist diagnosis and intervention. The development of machine learning methods for attention-based ASD screening shows promises, yet it has been limited by the need for high-precision eye trackers, the scope of stimuli, and black-box neural networks, making it impractical for real-life clinical scenarios. This study proposes an interpretable and generalizable framework for quantifying atypical attention in people with ASD. Our framework utilizes photos taken by participants with standard cameras to enable practical and flexible deployment in resource-constrained regions. With an emphasis on interpretability and trustworthiness, our method automates human-like diagnostic reasoning, associates photos with semantically plausible attention patterns, and provides clinical evidence to support ASD experts. We further evaluate models on both in-domain and out-of-domain data and demonstrate that our approach accurately classifies individuals with ASD and generalizes across different domains. The proposed method offers an innovative, reliable, and cost-effective tool to assist the diagnostic procedure, which can be an important effort toward transforming clinical research in ASD screening with artificial intelligence systems. Our code is publicly available at \u0000<uri>https://github.com/szzexpoi/proto_asd</uri>\u0000.","PeriodicalId":54300,"journal":{"name":"IEEE Transactions on Cognitive and Developmental Systems","volume":"16 5","pages":"1803-1813"},"PeriodicalIF":5.0,"publicationDate":"2024-04-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140593962","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 : 2024-04-08DOI: 10.1109/TCDS.2024.3385849
Shan Zhong;Guoqiang Li;Wenhao Ying;Fuzhou Zhao;Gengsheng Xie;Shengrong Gong
Video object segmentation (VOS) uses the first annotated video mask to achieve consistent and precise segmentation in subsequent frames. Recently, memory-based methods have received significant attention owing to their substantial performance enhancements. However, these approaches rely on a fixed global memory strategy, which poses a challenge to segmentation accuracy and speed in the context of longer videos. To alleviate this limitation, we propose a novel semisupervised VOS model, founded on the principles of the adaptive memory network. Our proposed model adaptively extracts object features by focusing on the object area while effectively filtering out extraneous background noise. An identification mechanism is also thoughtfully applied to discern each object in multiobject scenarios. To further reduce storage consumption without compromising the saliency of object information, the outdated features residing in the memory pool are compressed into salient features through the employment of a self-attention mechanism. Furthermore, we introduce a local matching module, specifically devised to refine object features by fusing the contextual information from historical frames. We demonstrate the efficiency of our approach through experiments, substantially augmenting both the speed and precision of segmentation for long-term videos, while maintaining comparable performance for short videos.
视频对象分割（VOS）利用首次注释的视频掩码来实现后续帧的一致和精确分割。最近，基于内存的方法因其性能大幅提升而备受关注。然而，这些方法依赖于固定的全局内存策略，这对较长视频的分割精度和速度提出了挑战。为了缓解这一限制，我们根据自适应记忆网络的原理，提出了一种新颖的半监督 VOS 模型。我们提出的模型通过聚焦对象区域自适应地提取对象特征，同时有效地过滤掉无关的背景噪声。此外，我们还贴心地采用了一种识别机制，以辨别多物体场景中的每个物体。为了在不影响对象信息显著性的前提下进一步减少存储消耗，我们采用了自我关注机制，将内存池中的过时特征压缩成显著特征。此外，我们还引入了一个局部匹配模块，专门用于通过融合历史帧的上下文信息来完善对象特征。我们通过实验证明了这一方法的高效性，大大提高了长期视频的分割速度和精度，同时保持了与短视频相当的性能。
{"title":"Efficient Semisupervised Object Segmentation for Long-Term Videos Using Adaptive Memory Network","authors":"Shan Zhong;Guoqiang Li;Wenhao Ying;Fuzhou Zhao;Gengsheng Xie;Shengrong Gong","doi":"10.1109/TCDS.2024.3385849","DOIUrl":"10.1109/TCDS.2024.3385849","url":null,"abstract":"Video object segmentation (VOS) uses the first annotated video mask to achieve consistent and precise segmentation in subsequent frames. Recently, memory-based methods have received significant attention owing to their substantial performance enhancements. However, these approaches rely on a fixed global memory strategy, which poses a challenge to segmentation accuracy and speed in the context of longer videos. To alleviate this limitation, we propose a novel semisupervised VOS model, founded on the principles of the adaptive memory network. Our proposed model adaptively extracts object features by focusing on the object area while effectively filtering out extraneous background noise. An identification mechanism is also thoughtfully applied to discern each object in multiobject scenarios. To further reduce storage consumption without compromising the saliency of object information, the outdated features residing in the memory pool are compressed into salient features through the employment of a self-attention mechanism. Furthermore, we introduce a local matching module, specifically devised to refine object features by fusing the contextual information from historical frames. We demonstrate the efficiency of our approach through experiments, substantially augmenting both the speed and precision of segmentation for long-term videos, while maintaining comparable performance for short videos.","PeriodicalId":54300,"journal":{"name":"IEEE Transactions on Cognitive and Developmental Systems","volume":"16 5","pages":"1789-1802"},"PeriodicalIF":5.0,"publicationDate":"2024-04-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140593831","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 : 2024-04-08DOI: 10.1109/TCDS.2024.3386364
Md. Nurul Ahad Tawhid;Siuly Siuly;Kate Wang;Hua Wang
The global health burden of neurological disorders (NDs) is vast, and they are recognized as major causes of mortality and disability worldwide. Most existing NDs detection methods are disease-specific, which limits an algorithm's cross-disease applicability. A single diagnostic platform can save time and money over multiple diagnostic systems. There is currently no unified standard platform for diagnosing different types of NDs utilizing electroencephalogram (EEG) signal data. To address this issue, this study aims to develop a generic EEG neural Network (GENet) framework based on a convolutional neural network that can identify various NDs from EEG. The proposed framework consists of several parts: 1) preparing data using channel reduction, resampling, and segmentation for the GENet model; 2) designing and training the GENet model to carry out important features for the classification task; and 3) assessing the proposed model's performance using different signal segment lengths and several training batch sizes and also cross-validating using seven different EEG datasets of six distinct NDs namely schizophrenia, autism spectrum disorder, epilepsy, Parkinson's disease, mild cognitive impairment, and attention-deficit/hyperactivity disorder. In addition, this study also investigates whether the proposed GENet model can identify multiple NDs from EEG. The proposed model achieved much better performance for both binary and multiclass classification compared to state-of-the-art methods. In addition, the proposed model is validated using several ablation studies and layerwise feature visualization, which provide consistency and efficiency to the proposed model. The proposed GENet model will help technologists create standard software for detecting any of these NDs from EEG.
{"title":"GENet: A Generic Neural Network for Detecting Various Neurological Disorders From EEG","authors":"Md. Nurul Ahad Tawhid;Siuly Siuly;Kate Wang;Hua Wang","doi":"10.1109/TCDS.2024.3386364","DOIUrl":"10.1109/TCDS.2024.3386364","url":null,"abstract":"The global health burden of neurological disorders (NDs) is vast, and they are recognized as major causes of mortality and disability worldwide. Most existing NDs detection methods are disease-specific, which limits an algorithm's cross-disease applicability. A single diagnostic platform can save time and money over multiple diagnostic systems. There is currently no unified standard platform for diagnosing different types of NDs utilizing electroencephalogram (EEG) signal data. To address this issue, this study aims to develop a generic EEG neural Network (GENet) framework based on a convolutional neural network that can identify various NDs from EEG. The proposed framework consists of several parts: 1) preparing data using channel reduction, resampling, and segmentation for the GENet model; 2) designing and training the GENet model to carry out important features for the classification task; and 3) assessing the proposed model's performance using different signal segment lengths and several training batch sizes and also cross-validating using seven different EEG datasets of six distinct NDs namely schizophrenia, autism spectrum disorder, epilepsy, Parkinson's disease, mild cognitive impairment, and attention-deficit/hyperactivity disorder. In addition, this study also investigates whether the proposed GENet model can identify multiple NDs from EEG. The proposed model achieved much better performance for both binary and multiclass classification compared to state-of-the-art methods. In addition, the proposed model is validated using several ablation studies and layerwise feature visualization, which provide consistency and efficiency to the proposed model. The proposed GENet model will help technologists create standard software for detecting any of these NDs from EEG.","PeriodicalId":54300,"journal":{"name":"IEEE Transactions on Cognitive and Developmental Systems","volume":"16 5","pages":"1829-1842"},"PeriodicalIF":5.0,"publicationDate":"2024-04-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140593957","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 : 2024-04-04DOI: 10.1109/TCDS.2024.3373151
{"title":"IEEE Transactions on Cognitive and Developmental Systems Publication Information","authors":"","doi":"10.1109/TCDS.2024.3373151","DOIUrl":"https://doi.org/10.1109/TCDS.2024.3373151","url":null,"abstract":"","PeriodicalId":54300,"journal":{"name":"IEEE Transactions on Cognitive and Developmental Systems","volume":"16 2","pages":"C2-C2"},"PeriodicalIF":5.0,"publicationDate":"2024-04-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10491580","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140348382","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-04-04DOI: 10.1109/TCDS.2024.3372274
Junpei Zhong;Ran Dong;Soichiro Ikuno;Yanan Li;Chenguang Yang
Movements play a critical role in robotic systems, with considerations varying across different robotic systems regarding factors, such as accuracy, speed, energy consumption, and naturalness of movements in various parts of the robotic mechanics. Over the past decades, the robotics community has developed computationally efficient mathematical tools for studying, simulating, and optimizing movements of articulated bodies to address these challenges.
{"title":"Guest Editorial Special Issue on Movement Sciences in Cognitive Systems","authors":"Junpei Zhong;Ran Dong;Soichiro Ikuno;Yanan Li;Chenguang Yang","doi":"10.1109/TCDS.2024.3372274","DOIUrl":"https://doi.org/10.1109/TCDS.2024.3372274","url":null,"abstract":"Movements play a critical role in robotic systems, with considerations varying across different robotic systems regarding factors, such as accuracy, speed, energy consumption, and naturalness of movements in various parts of the robotic mechanics. Over the past decades, the robotics community has developed computationally efficient mathematical tools for studying, simulating, and optimizing movements of articulated bodies to address these challenges.","PeriodicalId":54300,"journal":{"name":"IEEE Transactions on Cognitive and Developmental Systems","volume":"16 2","pages":"403-406"},"PeriodicalIF":5.0,"publicationDate":"2024-04-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10491578","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140348383","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-04-04DOI: 10.1109/TCDS.2024.3373155
{"title":"IEEE Transactions on Cognitive and Developmental Systems Information for Authors","authors":"","doi":"10.1109/TCDS.2024.3373155","DOIUrl":"https://doi.org/10.1109/TCDS.2024.3373155","url":null,"abstract":"","PeriodicalId":54300,"journal":{"name":"IEEE Transactions on Cognitive and Developmental Systems","volume":"16 2","pages":"C4-C4"},"PeriodicalIF":5.0,"publicationDate":"2024-04-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10491285","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140348385","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","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：