Query：
学者姓名：刘贵忠
Refining：
Year
Type
Indexed by
Source
Complex
Co-Author
Language
Clean All
Abstract ：
3D objection detection is a key task in autonomous driving. Because 3D structure information is lost during perspective projection, 3D localization of an object from monocular images is challenging. We herein present a monocular 3D object detection method that formulates the 3D object localization as a paired keypoints regression problem. Our method exploits 2D bounding box priors to predict the projection of paired 3D keypoints on the image plane for each object, and the object localization is recovered via an inverse projection. A fast keypoint regression network is proposed to predict the projection of keypoints and to generate the initial 3D bounding box. Furthermore, to obtain more accurate 3D detection results, we leverage a light-weight cascaded refinement module to rectify the initial 3D box, which takes the instance point cloud converted from the monocular depth prediction as input. Experiments on the KITTI dataset demonstrate that our method exhibits state-of-the-art performance solely via monocular images. Our method achieves 15.97, 10.42, and 7.91 3D AP on the three difficulty levels on the KITTI test set, respectively.
Keyword ：
3D Object Detection Deep Learning Instance Point Cloud Keypoint Detection
Cite：
Copy from the list or Export to your reference management。
| GB/T 7714 | Ji, Chaofeng , Liu, Guizhong , Zhao, Dan . Monocular 3D object detection via estimation of paired keypoints for autonomous driving [J]. | MULTIMEDIA TOOLS AND APPLICATIONS , 2022 , 81 (4) : 5973-5988 . |
| MLA | Ji, Chaofeng 等. "Monocular 3D object detection via estimation of paired keypoints for autonomous driving" . | MULTIMEDIA TOOLS AND APPLICATIONS 81 . 4 (2022) : 5973-5988 . |
| APA | Ji, Chaofeng , Liu, Guizhong , Zhao, Dan . Monocular 3D object detection via estimation of paired keypoints for autonomous driving . | MULTIMEDIA TOOLS AND APPLICATIONS , 2022 , 81 (4) , 5973-5988 . |
| Export to | NoteExpress RIS BibTex |
Abstract ：
Object tracking is a fundamental problem of computer vision. Although being studied for decades, the single object tracking problem has not been completely solved, since there exist various challenges in the real physical world, such as object deformation, complex background and imperfect imaging, which make tracking difficult. For these challenges, we design a robust feature extraction network. Specifically, we propose a novel channel-wise feature attention mechanism, which is integrated into the pipeline of a well-known convolutional neural network based visual tracking algorithm. It is crucial to represent the object robustly. Due to the representative feature, the tracking performance is improved. In experiments, we test the proposed tracking algorithm in OTB100, VOT2018, VOT2020 and VOT-TIR datasets. Compared to the baseline algorithm, our proposed algorithm obtains consistent performance improvement for different benchmarks with absolute increase of tracking success score in OTB100 up to 0.6, and absolute increase of EAO up to 0.022, 0.007, and 0.008 in VOT2018, VOT2020, VOT-TIR2015 respectively. The source codes are publicly available.
Keyword ：
Deep learning Feature attention Object tracking Representation learning
Cite：
Copy from the list or Export to your reference management。
| GB/T 7714 | Wu, Han , Liu, Guizhong . Split-merge-excitation: a robust channel-wise feature attention mechanism applied to MDNet tracking [J]. | MULTIMEDIA TOOLS AND APPLICATIONS , 2022 , 81 (28) : 40737-40754 . |
| MLA | Wu, Han 等. "Split-merge-excitation: a robust channel-wise feature attention mechanism applied to MDNet tracking" . | MULTIMEDIA TOOLS AND APPLICATIONS 81 . 28 (2022) : 40737-40754 . |
| APA | Wu, Han , Liu, Guizhong . Split-merge-excitation: a robust channel-wise feature attention mechanism applied to MDNet tracking . | MULTIMEDIA TOOLS AND APPLICATIONS , 2022 , 81 (28) , 40737-40754 . |
| Export to | NoteExpress RIS BibTex |
Abstract ：
We present a novel and high-performance framework for 3D object detection using stereo vision. This framework incorporates direct instance depth estimation efficiently, improving the accuracy of the final 3D object detection. Instead of detecting objects separately in the left and right images of a stereo display, we exploit a modified 2D object detector that takes only the left image as input to generate union 2D bounding boxes for both images, and to predict the depth of the 3D box center for each object. Using the union 2D boxes, we propose a direct instance-level depth estimation network, taking the estimated depth as guidance, to predict the desired depths for pixels belonging to an object from a small search range. This approach greatly improves the efficiency and accuracy of 3D detection. Moreover, we design an adaptive spatial feature aggregation module that can weaken the effect of background points and automatically integrate important instance features to achieve accurate 3D object localization. Our method outperforms current state-of-the-art stereo-based 3D detection methods on the KITTI benchmark dataset, and it can efficiently employ a shared model for 3D multi-class detection. Code will be available at https://github.com/xjtuwh/iDepNet/trree/master.
Keyword ：
3D object detection Autonomous driving Deep learning Instance depth estimation Stereo images
Cite：
Copy from the list or Export to your reference management。
| GB/T 7714 | Ji, Chaofeng , Liu, Guizhong , Zhao, Dan . Stereo 3D object detection via instance depth prior guidance and adaptive spatial feature aggregation [J]. | VISUAL COMPUTER , 2022 . |
| MLA | Ji, Chaofeng 等. "Stereo 3D object detection via instance depth prior guidance and adaptive spatial feature aggregation" . | VISUAL COMPUTER (2022) . |
| APA | Ji, Chaofeng , Liu, Guizhong , Zhao, Dan . Stereo 3D object detection via instance depth prior guidance and adaptive spatial feature aggregation . | VISUAL COMPUTER , 2022 . |
| Export to | NoteExpress RIS BibTex |
Abstract ：
Knowledge Graph Question Answering (KGQA) aims to answer user-questions from a knowledge graph (KG) by identifying the reasoning relations between topic entity and answer. As a complex branch task of KGQA, multi-hop KGQA requires reasoning over the multi-hop relational chain preserved in KG to arrive at the right answer. Despite recent successes, the existing works on answering multi-hop complex questions still face the following challenges: (i) The absence of an explicit relational chain order reflected in user-question stems from a misunderstanding of a user's intentions. (ii) Incorrectly capturing relational types on weak supervision of which dataset lacks intermediate reasoning chain annotations due to expensive labeling cost. (iii) Failing to consider implicit relations between the topic entity and the answer implied in structured KG because of limited neighborhoods size constraint in subgraph retrieval-based algorithms. To address these issues in multi-hop KGQA, we propose a novel model herein, namely Relational Chain based Embedded KGQA (Rce-KGQA), which simultaneously utilizes the explicit relational chain revealed in natural language question and the implicit relational chain stored in structured KG. Our extensive empirical study on three open-domain benchmarks proves that our method significantly outperforms the state-of-the-art counterparts like GraftNet, PullNet and EmbedKGQA. Comprehensive ablation experiments also verify the effectiveness of our method on themulti-hop KGQA task. We havemade our model's source code available at github: https://github.com/ albert-jin/Rce-KGQA.
Keyword ：
Data mining and search Knowledge graph based multi-hop QA Knowledge graph embedding Neural semantic parsing Question answering
Cite：
Copy from the list or Export to your reference management。
| GB/T 7714 | Jin, Weiqiang , Zhao, Biao , Yu, Hang et al. Improving embedded knowledge graph multi-hop question answering by introducing relational chain reasoning [J]. | DATA MINING AND KNOWLEDGE DISCOVERY , 2022 , 37 (1) : 255-288 . |
| MLA | Jin, Weiqiang et al. "Improving embedded knowledge graph multi-hop question answering by introducing relational chain reasoning" . | DATA MINING AND KNOWLEDGE DISCOVERY 37 . 1 (2022) : 255-288 . |
| APA | Jin, Weiqiang , Zhao, Biao , Yu, Hang , Tao, Xi , Yin, Ruiping , Liu, Guizhong . Improving embedded knowledge graph multi-hop question answering by introducing relational chain reasoning . | DATA MINING AND KNOWLEDGE DISCOVERY , 2022 , 37 (1) , 255-288 . |
| Export to | NoteExpress RIS BibTex |
Abstract ：
Mobile edge computing (MEC) has become an indispensable part of the era of the intelligent manufacturing industry 4.0. In the smart city, computation-intensive tasks can be offloaded to the MEC server or the central cloud server for execution. However, the privacy disclosure issue may arise when the raw data is migrated to other MEC servers or the central cloud server. Since federated learning has the characteristics of protecting the privacy and improving training performance, it is introduced to solve the issue. In this article, we formulate the joint optimization problem of task offloading and resource allocation to minimize the energy consumption of all Internet of Things (IoT) devices subject to delay threshold and limited resources. A two-timescale federated deep reinforcement learning algorithm based on Deep Deterministic Policy Gradient (DDPG) framework (FL-DDPG) is proposed. Simulation results show that the proposed algorithm can greatly reduce the energy consumption of all IoT devices.
Keyword ：
DDPG federated learning mobile edge computing resource allocation smart city task offloading
Cite：
Copy from the list or Export to your reference management。
| GB/T 7714 | Chen, Xing , Liu, Guizhong . Federated Deep Reinforcement Learning-Based Task Offloading and Resource Allocation for Smart Cities in a Mobile Edge Network [J]. | SENSORS , 2022 , 22 (13) . |
| MLA | Chen, Xing et al. "Federated Deep Reinforcement Learning-Based Task Offloading and Resource Allocation for Smart Cities in a Mobile Edge Network" . | SENSORS 22 . 13 (2022) . |
| APA | Chen, Xing , Liu, Guizhong . Federated Deep Reinforcement Learning-Based Task Offloading and Resource Allocation for Smart Cities in a Mobile Edge Network . | SENSORS , 2022 , 22 (13) . |
| Export to | NoteExpress RIS BibTex |
Abstract ：
Discriminative correlation filters (DCF) have drawn increasing interest in visual tracking. In particular, a few recent works treat DCF as a special layer and add it into a Siamese network for visual tracking. However, most of them adopt shallow networks to learn target representations, which lack robust semantic information of deeper layers and make these works fail to handle significant appearance changes. In this paper, we design a novel Siamese network to fuse high-level semantic features and low-level spatial detail features for correlation tracking. Specifically, to introduce more semantic information into low-level features, we specially design a residual semantic embedding module to adaptively involve more semantic information from high-level features to guide the feature fusion. Furthermore, we adopt an effective and efficient channel attention mechanism to filter out noise information and make the network focus more on valuable features that are beneficial for visual tracking. The overall architecture is trained end-to-end offline to adaptively learn target representations, which are not only enabled to encode high-level semantic features and low-level spatial detail features, but also closely related to correlation filters. Experimental results on widely used OTB2013, OTB2015, VOT2016, TC-128, and UAV123 benchmarks show that our proposed tracker performs favorably against several state-of-the-art trackers.
Keyword ：
Correlation correlation filters deep features Feature extraction Fuses Information filters multi-level feature fusion Semantics Target tracking Visualization Visual tracking
Cite：
Copy from the list or Export to your reference management。
| GB/T 7714 | Liu, Guangen , Liu, Guizhong . End-to-End Correlation Tracking With Enhanced Multi-Level Feature Fusion [J]. | IEEE ACCESS , 2021 , 9 : 128827-128840 . |
| MLA | Liu, Guangen et al. "End-to-End Correlation Tracking With Enhanced Multi-Level Feature Fusion" . | IEEE ACCESS 9 (2021) : 128827-128840 . |
| APA | Liu, Guangen , Liu, Guizhong . End-to-End Correlation Tracking With Enhanced Multi-Level Feature Fusion . | IEEE ACCESS , 2021 , 9 , 128827-128840 . |
| Export to | NoteExpress RIS BibTex |
Abstract ：
Recognition of small moving targets in space has become one of the frontier scientific researches in recent decade. Most of them focus on detection and recognition in star image with sidereal stare mode. However, in this research field, few researches are about detection and recognition in star image with track rate mode. In this paper, a novel approach is proposed to recognize the moving target in single frame by machine learning method based on elliptical characteristic extraction of star points. The technical path about recognition of moving target in space is redesigned instead of traditional processing approaches. Elliptical characteristics of each star point can be successfully extracted from single image. Machine learning can achieve the classification goal in order to make sure that all moving targets can be extracted. The experiments show that our proposed approach can have better performance in star images with different qualities. © 2021 World Scientific Publishing Company.
Keyword ：
Machine learning Stars
Cite：
Copy from the list or Export to your reference management。
| GB/T 7714 | Du, Yun , Wen, Desheng , Liu, Guizhong et al. Small Moving Target Recognition in Star Image with TRM [J]. | International Journal of Pattern Recognition and Artificial Intelligence , 2021 , 35 (2) . |
| MLA | Du, Yun et al. "Small Moving Target Recognition in Star Image with TRM" . | International Journal of Pattern Recognition and Artificial Intelligence 35 . 2 (2021) . |
| APA | Du, Yun , Wen, Desheng , Liu, Guizhong , Qiu, Shi . Small Moving Target Recognition in Star Image with TRM . | International Journal of Pattern Recognition and Artificial Intelligence , 2021 , 35 (2) . |
| Export to | NoteExpress RIS BibTex |
Abstract ：
The self-supervised depth and camera pose estimation methods are proposed to address the difficulty of acquiring the densely labeled ground-truth data and have achieved a great advance. As the stereo vision could constrain the predicted depth to a real-world scale, in this paper, we study the use of both left-right pairs and adjacent frames of stereo sequences for self-supervised semantic and optical flow guided monocular depth and camera pose estimation without real pose information. In particular, we explore (i) to construct a cascaded structure of the depth-pose and optical flow for well-initializing the optical flow, (ii) a cycle learning strategy to further constrain the depth-pose learning by the cross-task consistency, and (iii) a weighted semantic guided smoothness loss to match the real nature of a depth map. Our method produces favorable results against the state-of-the-art methods on several benchmarks. And we also demonstrate the generalization ability of our method on the cross dataset.
Keyword ：
Camera pose estimation Monocular depth estimation Self-supervised learning Stereo vision
Cite：
Copy from the list or Export to your reference management。
| GB/T 7714 | Fang, Jiaojiao , Liu, Guizhong . Semantic and Optical Flow Guided Self-supervised Monocular Depth and Ego-Motion Estimation [J]. | IMAGE AND GRAPHICS (ICIG 2021), PT III , 2021 , 12890 : 465-477 . |
| MLA | Fang, Jiaojiao et al. "Semantic and Optical Flow Guided Self-supervised Monocular Depth and Ego-Motion Estimation" . | IMAGE AND GRAPHICS (ICIG 2021), PT III 12890 (2021) : 465-477 . |
| APA | Fang, Jiaojiao , Liu, Guizhong . Semantic and Optical Flow Guided Self-supervised Monocular Depth and Ego-Motion Estimation . | IMAGE AND GRAPHICS (ICIG 2021), PT III , 2021 , 12890 , 465-477 . |
| Export to | NoteExpress RIS BibTex |
Abstract ：
The quadrant detector (QD), has developed into a core detector in the free space optical communication system. The light power received by the detector surface will be very weak after long distance transmission of laser, it brings great challenges to the high precision spot position detection of the detector. Therefore, this letter proposes a method to improve the spot position detection accuracy of the QD through artificial neural network. The neural network can solve the impact of multiple different factors on the detection accuracy of the detector at one time, which can save a lot of time and cost. Moreover, the test results of the detection accuracy of the network show that the neural network has significantly improved the detection accuracy of the spot position of the QD.
Keyword ：
Detectors Free space optical communication (FSOC) Laser beams neural network Neural networks Optical fiber amplifiers Position measurement quadrant detector (QD) Surface treatment Training
Cite：
Copy from the list or Export to your reference management。
| GB/T 7714 | Wang, Xuan , Su, Xiuqin , Liu, Guizhong et al. Method to Improve the Detection Accuracy of Quadrant Detector Based on Neural Network [J]. | IEEE PHOTONICS TECHNOLOGY LETTERS , 2021 , 33 (22) : 1254-1257 . |
| MLA | Wang, Xuan et al. "Method to Improve the Detection Accuracy of Quadrant Detector Based on Neural Network" . | IEEE PHOTONICS TECHNOLOGY LETTERS 33 . 22 (2021) : 1254-1257 . |
| APA | Wang, Xuan , Su, Xiuqin , Liu, Guizhong , Han, Junfeng , Zhu, Wenhua , Liu, Zengxin . Method to Improve the Detection Accuracy of Quadrant Detector Based on Neural Network . | IEEE PHOTONICS TECHNOLOGY LETTERS , 2021 , 33 (22) , 1254-1257 . |
| Export to | NoteExpress RIS BibTex |
Abstract ：
The augmented reality (AR) applications have been widely used in the field of Internet of Things (IoT) because of good immersion experience for users, but their ultralow delay demand and high energy consumption bring a huge challenge to the current communication system and terminal power. The emergence of mobile-edge computing (MEC) provides a good thinking to solve this challenge. In this article, we study an energy-efficient task offloading and resource allocation scheme for AR in both the single-MEC and multi-MEC systems. First, a more specific and detailed AR application model is established as a directed acyclic graph according to its internal functionality. Second, based on this AR model, a joint optimization problem of task offloading and resource allocation is formulated to minimize the energy consumption of each user subject to the latency requirement and the limited resources. The problem is a mixed multiuser competition and cooperation problem, which involves the task offloading decision, uplink/downlink transmission resources allocation, and computing resources allocation of users and MEC server. Since it is an NP-hard problem and the communication environment is dynamic, it is difficult for genetic algorithms or heuristic algorithms to solve. Therefore, we propose an intelligent and efficient resource allocation and task offloading algorithm based on the deep reinforcement learning framework of multiagent deep deterministic policy gradient (MADDPG) in a dynamic communication environment. Finally, simulation results show that the proposed algorithm can greatly reduce the energy consumption of each user terminal.
Keyword ：
Augmented reality (AR) Computational modeling deep reinforcement learning Energy consumption Heuristic algorithms Internet of Things (IoT) mobile-edge computing (MEC) multiagent deep deterministic policy gradient (MADDPG) Optimization resource allocation Resource management Servers Task analysis task offloading
Cite：
Copy from the list or Export to your reference management。
| GB/T 7714 | Chen, Xing , Liu, Guizhong . Energy-Efficient Task Offloading and Resource Allocation via Deep Reinforcement Learning for Augmented Reality in Mobile Edge Networks [J]. | IEEE INTERNET OF THINGS JOURNAL , 2021 , 8 (13) : 10843-10856 . |
| MLA | Chen, Xing et al. "Energy-Efficient Task Offloading and Resource Allocation via Deep Reinforcement Learning for Augmented Reality in Mobile Edge Networks" . | IEEE INTERNET OF THINGS JOURNAL 8 . 13 (2021) : 10843-10856 . |
| APA | Chen, Xing , Liu, Guizhong . Energy-Efficient Task Offloading and Resource Allocation via Deep Reinforcement Learning for Augmented Reality in Mobile Edge Networks . | IEEE INTERNET OF THINGS JOURNAL , 2021 , 8 (13) , 10843-10856 . |
| Export to | NoteExpress RIS BibTex |
Export
| Results： |
Selected to |
| Format： |
