Article
Automation & Control Systems
Sheng Yu, Di-Hua Zhai, Yuanqing Xia
Summary: This paper proposes a new robotic grasp detection network called SKGNet, which achieves high accuracy and real-time performance by integrating attention mechanism and multi-scale fusion features. Training and testing on public datasets, SKGNet achieves superior accuracy and high detection speed compared to existing methods. Real-world experiments demonstrate the generalization performance and effectiveness of SKGNet.
IEEE TRANSACTIONS ON AUTOMATION SCIENCE AND ENGINEERING
(2022)
Article
Chemistry, Analytical
Shervin Minaee, Mehdi Minaei, Amirali Abdolrashidi
Summary: Facial expression recognition is an active research area with challenges in high intra-class variation. Deep learning models have shown better performance, but there is still room for improvement. This work proposes a deep learning approach based on an attentional convolutional network, achieving significant improvements on multiple datasets.
Article
Computer Science, Information Systems
Jun Ho Chung, Dong Won Kim, Tae Koo Kang, Myo Taeg Lim
Summary: Convolutional Neural Networks (CNNs) have become primary technologies in computer vision systems. This paper proposes an attentional-deconvolution module (ADM)-based net(ADM-Net) to improve classification under noise-coupled environments. ADM-Net achieves the highest records in different noise cases, demonstrating its robustness effectively.
MULTIMEDIA TOOLS AND APPLICATIONS
(2022)
Article
Chemistry, Analytical
Bohan Liu, Ruixing Ge, Yuxuan Zhu, Bolin Zhang, Xiaokai Zhang, Yanfei Bao
Summary: This paper introduces a multi-modal fusion method called iDAF, which integrates multi-modal data through iterative dual-scale attentional fusion. The method not only extracts recognition characteristics from specific domains, but also complements the strengths of different modalities, achieving excellent performance.
Article
Engineering, Electrical & Electronic
Yuanhao Li, Yu Liu, Zhiqiang Ma, Panfeng Huang
Summary: In this article, a novel generative convolutional neural network model is proposed to improve the accuracy and robustness of robot grasp detection in real-world scenes. The method achieves excellent performance on the Cornell Grasping datasets and Jacquard datasets, and is put to the test in a real-world robot grasping scenario.
IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT
(2022)
Article
Automation & Control Systems
Dexin Wang, Chunsheng Liu, Faliang Chang, Nanjun Li, Guangxin Li
Summary: This article presents a pixel-level grasp detection method based on deep neural network, which achieves high accuracy and fast speed on RGB images through novel representation models, adaptive attribute models, and feature fusion and grasp-aware network.
IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS
(2022)
Article
Automation & Control Systems
Hu Cao, Guang Chen, Zhijun Li, Qian Feng, Jianjie Lin, Alois Knoll
Summary: This study proposes an efficient grasp detection network for robotic grasp tasks. The network uses a lightweight generative structure to achieve a balance between high grasp confidence and fast inference speed. It introduces a Gaussian kernel-based grasp representation for encoding training samples and employs receptive field blocks and attention mechanisms for improved feature discriminability and semantic information fusion. Experimental results demonstrate excellent performance on Cornell, Jacquard, and extended OCID grasp datasets with accuracy of 97.8%, 95.6%, and 76.4% respectively.
IEEE-ASME TRANSACTIONS ON MECHATRONICS
(2023)
Article
Computer Science, Interdisciplinary Applications
Yongjian Li, Liting Zhang, Lin Zhu, Lei Liu, Baokun Han, Yatao Zhang, Shoushui Wei
Summary: In this study, a self-complementary attentional convolutional neural network (SC-CNN) was designed to accurately identify atrial fibrillation (AF) in wearable dynamic electrocardiographic (ECG) signals. The proposed method achieved high accuracies and AUC values on three public databases, as well as a high sensitivity on the clinical database, demonstrating its effectiveness in AF diagnosis.
COMPUTER METHODS AND PROGRAMS IN BIOMEDICINE
(2023)
Article
Multidisciplinary Sciences
Sukhendra Singh, Manoj Kumar, Abhay Kumar, Birendra Kumar Verma, S. Shitharth
Summary: Worldwide, pneumonia is a leading cause of infant mortality. Chest X-rays are used by experienced radiologists to diagnose pneumonia and other respiratory diseases, but the complexity of the diagnostic procedure often leads to disagreement among radiologists. Early diagnosis is crucial for mitigating the impact of the disease, and computer-aided diagnostics, such as Quaternion neural networks, improve the accuracy of diagnosis. Our proposed QCSA network combines spatial and channel attention mechanisms with Quaternion residual network to classify chest X-Ray images for Pneumonia detection, achieving a high accuracy rate of 94.53% and an AUC of 0.89. This approach shows promising potential for detecting pneumonia.
SCIENTIFIC REPORTS
(2023)
Article
Computer Science, Artificial Intelligence
Getachew Ambaye, Enkhsaikhan Boldsaikhan, Krishna Krishnan
Summary: This study uses finite element simulation and machine learning to investigate the influence of cracks on the vibration characteristics of robot links. By transforming the vibration response data into image data and training a convolutional neural network, crack detection and analysis can be achieved.
NEURAL COMPUTING & APPLICATIONS
(2023)
Article
Computer Science, Artificial Intelligence
Yibo Gao, Huan Wang, Zuhao Liu
Summary: In this study, a residual-based temporal attention convolutional neural network (RTA-CNN) was proposed for atrial fibrillation (AF) detection, which automatically focuses on parts with more semantic information to achieve better performance. Additionally, a novel exponential nonlinearity loss (EN-Loss) was introduced to address the imbalance problem.
KNOWLEDGE-BASED SYSTEMS
(2021)
Article
Biology
Amir Ebrahimi, Suhuai Luo, Raymond Chiong
Summary: The study examined the effectiveness of applying deep sequence-based network models for AD detection, addressing the classification accuracy issue of 2D and 3D CNNs in AD detection by handling the MRI feature sequences generated by CNNs.
COMPUTERS IN BIOLOGY AND MEDICINE
(2021)
Article
Computer Science, Artificial Intelligence
Chao Yang, Xianzhi Wang, Lina Yao, Guodong Long, Jing Jiang, Guandong Xu
Summary: Multivariate time series classification is a critical problem in data mining with broad applications. We propose a novel convolutional neural network architecture called Attentional Gated Res2Net for multivariate time series classification. Our model utilizes hierarchical residual-like connections to capture multi-granular temporal information and two types of attention modules for better leveraging the temporal patterns. Experimental results demonstrate that our model outperforms several baselines and state-of-the-art methods on 14 benchmark multivariate time-series datasets.
NEURAL PROCESSING LETTERS
(2023)
Article
Multidisciplinary Sciences
Di Lu, Shuli Cheng, Liejun Wang, Shiji Song
Summary: In this paper, a new deep learning-based method for change detection is proposed, which utilizes multi-scale feature fusion and distribution strategies to improve the accuracy of change region detection. Experimental results demonstrate that this method outperforms other comparative methods.
SCIENTIFIC REPORTS
(2022)
Article
Automation & Control Systems
Meng Xiao, Bo Yang, Shilong Wang, Zhengping Zhang, Yan He
Summary: This paper proposes a Fine Coordinate Attention (FCA) block to address the challenges of surface defect detection. The FCA block encodes both average and salient information in two coordinate directions, capturing spatial dependence and achieving long-range interaction. Experimental results show that the FCA block outperforms existing attention mechanisms in image classification and object detection tasks.
ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE
(2023)
Article
Automation & Control Systems
Sheng Yu, Di-Hua Zhai, Yuanqing Xia
Summary: In this article, a novel grasp detection network called efficient grasp detection network (EGNet) is proposed to address the challenges of grasping in stacked scenes. It combines object detection, grasp detection, and manipulation relationship detection tasks. The EGNet adopts the EfficientDet for object detection and modifies some hyperparameters. It introduces a novel grasp detection module that utilizes the feature map from bidirectional feature pyramid network (BiFPN) to output the grasp position and quality score. The EGNet also incorporates manipulation relation analysis and achieves high detection accuracy in different datasets and practical scenarios.
IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS
(2023)
Article
Automation & Control Systems
Sheng Yu, Di-Hua Zhai, Yuanqing Xia
Summary: This article proposes a novel robotic grasp detection method, CGNet, which can accurately detect grasp rectangles in cluttered scenes. By introducing attention module, grasp region proposal module, and position focal loss, the detection accuracy is improved. The experiments demonstrate the effectiveness of the proposed method on various datasets and in the real world.
IEEE-ASME TRANSACTIONS ON MECHATRONICS
(2023)
Article
Automation & Control Systems
Sheng Yu, Di-Hua Zhai, Yuanqing Xia
Summary: This paper proposes a new robotic grasp detection network called SKGNet, which achieves high accuracy and real-time performance by integrating attention mechanism and multi-scale fusion features. Training and testing on public datasets, SKGNet achieves superior accuracy and high detection speed compared to existing methods. Real-world experiments demonstrate the generalization performance and effectiveness of SKGNet.
IEEE TRANSACTIONS ON AUTOMATION SCIENCE AND ENGINEERING
(2022)
Article
Automation & Control Systems
Sheng Yu, Di-Hua Zhai, Yuanqing Xia
Summary: The paper proposes a new category-level object pose estimation network, SCNet, which introduces the prior point cloud and RGB images to estimate and refine the pose of target objects. Experimental results show that the proposed method outperforms other methods in object pose estimation and achieves good performance in robotic grasp tasks.
IEEE-ASME TRANSACTIONS ON MECHATRONICS
(2023)
Article
Automation & Control Systems
Di-Hua Zhai, Sheng Yu, Yuanqing Xia
Summary: This research proposes a network called FANet based on grasp keypoints, which achieves real-time and accurate robotic grasp detection. It includes a local refinement module, a global feature refinement module, and a grasp keypoint optimization module. Results show that FANet achieves excellent detection performance on multiple datasets and has a high success rate in real-world grasp experiments with a Baxter robot.
IEEE TRANSACTIONS ON AUTOMATION SCIENCE AND ENGINEERING
(2023)
Article
Computer Science, Artificial Intelligence
Sheng Yu, Di-Hua Zhai, Yuanqing Xia
Summary: This article proposes a robotic grasping network (GN) that combines pushing and grasping actions to solve the problem of grasping in cluttered scenes. The pushing action is performed by a vision transformer network (PTNet) to predict the object position, while the grasping detection is achieved by a cross dense fusion network (CDFNet) to accurately locate the optimal grasping position. Through simulations and actual robot experiments, the proposed network achieves state-of-the-art performance.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2023)
Proceedings Paper
Automation & Control Systems
Sheng Yu, Di-Hua Zhai, Haocun Wu, Hongda Yang, Yuanqing Xia
PROCEEDINGS OF THE 39TH CHINESE CONTROL CONFERENCE
(2020)