检索结果-内蒙古大学图书馆

Soft Computing 2024年第17-18期28卷 9575-9586页

作者： Zhou, Tongle Chen, Mou Wang, Yuhui Zhu, Ronggang Yang, Chenguang College of Automation Engineering Nanjing University of Aeronautics and Astronautics Nanjing211106 China Key Laboratory of Avionics System Integrated Technology Shanghai200241 China Science and Technology on Electro-optic Control Laboratory Luoyang Institute of Electro-Optical Equipment of Avic Luoyang471000 China Bristol Robotics Laboratory University of the West of England BristolBS16 1QY United Kingdom

Unmanned aerial vehicles (UAVs) have shown their superiority for applications in complicated military missions. A cooperative attack–defense decision-making method based on satisficing decision-enhanced wolf pack search (SDEWPS) algorithm is developed for multi-UAV air combat in this paper. Firstly, the multi-UAV air combat mathematical model is provided and the attack–defense decision-making constraints are defined. Besides the traditional air combat situation, the capability of UAVs and target information including target type and target intention are all considered in this paper to establish the air combat superiority function. Then, the wolf pack search (WPS) algorithm is used to solve the attack decision problem. To improve efficiency, the satisficing decision theory is employed to enhance the WPS to obtain the satisficing solution rather than optimal solution. The simulation results show that the developed method can realize the cooperative attack decision-making. © The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature 2024.

关键词： Decision making

A Continuous Illumination Estimator for Shadow Removal 13

学校读者我要写书评

暂无评论

A Continuous Illumination Estimator for Shadow Removal

13th IEEE International Conference on CYBER Technology in automation, Control, and Intelligent Systems, CYBER 2023

作者： Liu, Jiawei Fan, Huijie Su, Yiming Yan, Zhen Wang, Qiang Li, Wentao Tang, Yandong State Key Laboratory of Robotics Shenyang Institute of Automation Chinese Academy of Sciences Shenyang110016 China Institutes for Robotics and Intelligent Manufacturing Chinese Academy of Sciences Shenyang110169 China University of Chinese Academy of Sciences Beijing100049 China Key Laboratory of Manufacturing Industrial Integrated Shenyang University Shenyang110044 China

ISBN: (纸本)9798350315196

Shadow removal methods often encounter difficulties in completely removing shadows or avoiding excessively bright removal results in shadow areas. To address this challenge, we propose a continuous-intensity shadow synthesis model and an illumination estimator for shadow removal. Our shadow synthesis model generates visually realistic images with continuously varying shadow intensity to augment the current shadow removal dataset. By utilizing the synthesized shadow images, we train a network to accurately estimate the illumination intensity in shadow areas. The estimated illumination intensity is positive for a shadow input image and negative for an over-lit input image, thus our method can illuminate shadow areas or suppress over-lit areas. Experimental results demonstrate that our method outperforms current state-of-the-art shadow removal methods, particularly when handling shadows with a wide range of intensity variations. © 2023 IEEE.

关键词： Computer vision

Nested Deep Feature Attention Module for Underwater Image Enhancement 3

学校读者我要写书评

暂无评论

Nested Deep Feature Attention Module for Underwater Image En...

3rd International conference on Computer, Big Data and Artificial Intelligence, ICCBDAI 2022

作者： Xiao, Yeqing Wang, Qiang Li, Yupeng Tang, Yandong School of Computer Science and Engineering Shenyang Jianzhu University Shenyang110168 China Key Laboratory of Manufacturing Industrial Integrated Shenyang University Shenyang110044 China State Key Laboratory of Robotics Shenyang Institute of Automation Chinese Academy of Sciences Shenyang110016 China Institutes for Robotics and Intelligent Manufacturing Chinese Academy of Sciences Shenyang110169 China

The captured underwater images always suffer degradations because of absorption and light scattering in water. Thus, underwater image enhancement becomes indispensable as a precondition to carry out underwater tasks. Thus, we put forward an end-to-end structure to solve the problem of underwater image degradation. Our network uses continuously stacked deep-layer feature extraction modules to exploit more significant features. For purpose of enhancing the image contrast, we recover high-frequency information by hiring the nested residual attention groups in feature extraction modules, which can adaptively extract more deeper features. Moreover, channel attention module is hired in residual groups to emphasize important information. Comparing with other state-of-the-art approaches, our network performs best on quantitative and qualitative evaluations at datasets EUVP and UIEBD. © Published under licence by IOP Publishing Ltd.

关键词： Feature extraction

Diverse Representations Embedding for Lifelong Person Re-Identification

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Liu, Shiben Fan, Huijie Wang, Qiang Chen, Xiai Han, Zhi Tang, Yandong The State Key Laboratory of Robotics Shenyang Institute of Automation Chinese Academy of Sciences Shenyang110016 China The Institutes for Robotics and Intelligent Manufacturing Chinese Academy of Sciences Shenyang110169 China The University of Chinese Academy of Sciences Beijing100049 China The Institutes for Robotics and Intelligent Manufacturing Chinese Academy of Sciences Shenyang110016 China The Key Laboratory of Manufacturing Industrial Integrated Automation Shenyang University The State Key Laboratory of Robotics Shenyang Institute of Automation Chinese Academy of Sciences Shenyang110016 China

Lifelong Person Re-Identification (LReID) aims to continuously learn from successive data streams, matching individuals across multiple cameras. The key challenge for LReID is how to effectively preserve old knowledge while incrementally learning new information, which is caused by task-level domain gaps and limited old task datasets. Existing methods based on CNN backbone are insufficient to explore the representation of each instance from different perspectives, limiting model performance on limited old task datasets and new task datasets. Unlike these methods, we propose a Diverse Representations Embedding (DRE) framework that first explores a pure transformer for LReID. The proposed DRE preserves old knowledge while adapting to new information based on instance-level and task-level layout. Concretely, an Adaptive Constraint Module (ACM) is proposed to implement integration and push away operations between multiple overlapping representations generated by transformer-based backbone, obtaining rich and discriminative representations for each instance to improve adaptive ability of LReID. Based on the processed diverse representations, we propose Knowledge Update (KU) and Knowledge Preservation (KP) strategies at the task-level layout by introducing the adjustment model and the learner model. KU strategy enhances the adaptive learning ability of learner models for new information under the adjustment model prior, and KP strategy preserves old knowledge operated by representation-level alignment and logit-level supervision in limited old task datasets while guaranteeing the adaptive learning information capacity of the LReID model. Extensive experiments were conducted on eleven Re-ID datasets, including five seen datasets for training in order-1 and order-2 orders and six unseen datasets for inference. Compared to state-of-the-art methods, our method achieves significantly improved performance in holistic, large-scale, and occluded datasets. Our code will be availab

关键词： Embeddings

GMT: A Robust Global Association Model for Multi-Target Multi-Camera Tracking

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Fan, Huijie Zhao, Tinghui Wang, Qiang Fan, Baojie Tang, Yandong Liu, LianQing State Key Laboratory of Robotics Shenyang Institute of Automation Chinese Academy of Sciences Shenyang110016 China University of Chinese Academy of Sciences Beijing100049 China Key Laboratory of Manufacturing Industrial Integrated Shenyang University China Automation and AI College Nanjing University of Posts and Telecommunications Nanjing210049 China State Key Laboratory of Integrated Services Networks Xi’an710071 China

In the task of multi-target multi-camera (MTMC) tracking of pedestrians, the data association problem is a key issue and main challenge, especially with complications arising from camera movements, lighting variations, and obstructions. However, most MTMC models adopt two-step approaches, thus heavily depending on the results of the first-step tracking in practical applications. Moreover, the same targets crossing different cameras may exhibit significant appearance variations, which further increases the difficulty of cross-camera matching. To address the aforementioned issues, we propose a global online MTMC tracking model that addresses the dependency on the first tracking stage in two-step methods and enhances cross-camera matching. Specifically, we propose a transformer-based global MTMC association module to explore target associations across different cameras and frames, generating global trajectories directly. Additionally, to integrate the appearance and spatiotemporal features of targets, we propose a feature extraction and fusion module for MTMC tracking. This module enhances feature representation and establishes correlations between the features of targets across multiple cameras. To accommodate high scene diversity and complex lighting condition variations, we have established the VisionTrack dataset, which enables the development of models that are more generalized and robust to various environments. Our model demonstrates significant improvements over comparison methods on the VisionTrack dataset and others. Copyright © 2024, The Authors. All rights reserved.

关键词： Cameras

Distribution-aware Forgetting Compensation for Exemplar-Free Lifelong Person Re-identification

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Liu, Shiben Fan, Huijie Wang, Qiang Fan, Baojie Tang, Yandong Qu, Liangqiong State Key Laboratory of Robotics Shenyang Institute of Automation Chinese Academy of Sciences Shenyang110016 China University of Chinese Academy of Sciences Beijing100049 China Key Laboratory of Manufacturing Industrial Integrated Automation Shenyang University China Automation and AI College Nanjing University of Posts and Telecommunications Nanjing210049 China State Key Laboratory of Integrated Services Networks Xi’an710071 China Department of Statistics and Actuarial Science the Institute of Data Science The University of Hong Kong Hong Kong

Lifelong Person Re-identification (LReID) suffers from a key challenge in preserving old knowledge while adapting to new information. The existing solutions include rehearsal-based and rehearsal-free methods to address this challenge. Rehearsal-based approaches rely on knowledge distillation, continuously accumulating forgetting during the distillation process. Rehearsal-free methods insufficiently learn domain-specific distributions, leading to forgetfulness over time. To solve these issues, we propose a novel Distribution-aware Forgetting Compensation (DAFC) model that explores cross-domain shared representation learning, and domain-specific distribution awareness without using old exemplars or knowledge distillation. We propose a Text-driven Prompt Aggregation (TPA) that utilizes text features to enrich prompt components and guide the prompt model to learn fine-grained instance-level representations. This can enhance discriminative identity information and establish the foundation for domain distribution awareness. Then, Distribution-based Awareness and Integration (DAI) is designed to capture each domain-specific distribution by dedicated expert networks and adaptively consolidate them into a shared region in high-dimensional space. In this manner, DAI can consolidate and enhance cross-domain shared representation learning while alleviating catastrophic forgetting. Furthermore, we develop a Knowledge Consolidation Mechanism (KCM) that comprises instance-level discrimination and cross-domain consistency alignment strategies to facilitate model adaptive learning of new knowledge from the current domain and promote knowledge consolidation learning between acquired domain-specific distributions, respectively. Experimental results show that our DAFC outperforms state-of-the-art methods. Our code is available at https://***/LiuShiBen/ DAFC. Copyright © 2025, The Authors. All rights reserved.

关键词： Inference engines

A Continuous Illumination Estimator for Shadow Removal

学校读者我要写书评

暂无评论

A Continuous Illumination Estimator for Shadow Removal

IEEE International Conference on Cyber Technology in automation, Control, and Intelligent Systems

作者： Jiawei Liu Huijie Fan Yiming Su Zhen Yan Qiang Wang Wentao Li Yandong Tang State Key Laboratory of Robotics Shenyang Institute of Automation Chinese Academy of Sciences Shenyang China Institutes for Robotics and Intelligent Manufacturing Chinese Academy of Sciences Shenyang China University of Chinese Academy of Sciences Beijing China Key Laboratory of Manufacturing Industrial Integrated Shenyang University Shenyang China

关键词：

Domain Consistency Representation Learning for Lifelong Person Re-Identification

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Liu, Shiben Fan, Huijie Wang, Qiang Ren, Weihong Fan, Baojie Tang, Yandong The State Key Laboratory of Robotics Shenyang Institute of Automation Chinese Academy of Sciences Shenyang110016 China The University of Chinese Academy of Sciences Beijing100049 China The Key Laboratory of Manufacturing Industrial Integrated Shenyang University Shenyang110044 China The Harbin Institute of Technology Shenzhen518055 China The Department of Automation college Nanjing University of Posts and Telecommunications Nanjing210000 China

Lifelong person re-identification (LReID) exhibits a contradictory relationship between intra-domain discrimination and inter-domain gaps when learning from continuous data. Intra-domain discrimination focuses on individual nuances (i.e., clothing type, accessories, etc.), while inter-domain gaps emphasize domain consistency. Achieving a trade-off between maximizing intra-domain discrimination and minimizing inter-domain gaps is a crucial challenge for improving LReID performance. Most existing methods strive to reduce inter-domain gaps through knowledge distillation to maintain domain consistency. However, they often ignore intra-domain discrimination. To address this challenge, we propose a novel domain consistency representation learning (DCR) model that explores global and attribute-wise representations as a bridge to balance intra-domain discrimination and inter-domain gaps. At the intra-domain level, we explore the complementary relationship between global and attribute-wise representations to improve discrimination among similar identities. Excessive learning intra-domain discrimination can lead to catastrophic forgetting. We further develop an attribute-oriented anti-forgetting (AF) strategy that explores attribute-wise representations to enhance inter-domain consistency, and propose a knowledge consolidation (KC) strategy to facilitate knowledge transfer. Extensive experiments show that our DCR model achieves superior performance compared to state-of-the-art LReID methods. Our code will be available soon. © 2024, CC BY.

关键词： Contrastive Learning

Improve the Performance of Active Knee Prosthesis During Stairs Ascent by Using the Leading Reference Control System

学校读者我要写书评

暂无评论

Improve the Performance of Active Knee Prosthesis During Sta...

IEEE International Conference on Cyborg and Bionic Systems (CBS)

作者： Qinpeng Huang Wenhao Wei Bojun Huang Hui Wang Pingao Huang Guanglin Li School of Electronic Engineering and Automation Guilin Key Laboratory of Intelligence Integrated Automation in Guangxi Universities Guilin University of Electronic Technology Guangxi China CAS Key Laboratory of HumanMachine Intelligence-Synergy Systems Shenzhen Institutes of Advanced Technology (SIAT) Chinese Academy of Sciences (CAS) and the SIAT Branch Shenzhen Institute of Artificial Intelligence and Robotics for Society Shenzhen China

Strong coupling between the human and machine exists in many cyborg and bionic systems (CBS). Some input signals for control system barely change after the outputs work. In this situation, it is hard for the controller to calculate the required reference output of system. For example, in the stance phase of stairs ascent gait using an active knee prosthesis (AKP), most AKPs use the Inertial Measurement Unit (IMU) signal and the Ground Reaction Forces (GRF) signal to control the output of knee joint torque. However, these two signals hardly change unless the controlled object (joint torque) is already close to the actual requirement. Therefore, this study introduces a leading reference control system (LRCS) that uses an overshoot reference value to make the AKP more effective in providing positive energy in the abovementioned situation. In addition, the LRCS with different parameters was evaluated through the experiments of stairs ascent with AKP. Results show that LRCS with the suitable parameter can provide greater torque output when going upstairs, making the subject go upstairs easier.

关键词： Knee Couplings Torque Measurement units Inertial navigation Stairs Control systems

Image Inpainting via Enhanced Generative Adversarial Network

学校读者我要写书评

暂无评论

Image Inpainting via Enhanced Generative Adversarial Network

International Conference on Mechatronics and Machine Vision in Practice (M2VIP)

作者： Qiang Wang Huijie Fan Key Laboratory of Manufacturing Industrial Integrated Shenyang University the State Key Laboratory of Robotics Shenyang Institute of Automation Chinese Academy of Sciences Shenyang China State Key Laboratory of Robotics Shenyang Institute of Automation Chinese Academy of Sciences China China

ISBN: (纸本)9781665431545

In this paper, we propose an Enhanced Generative Model for Image Inpainting (EGMII). Unlike most state-of-the-art algorithms using extra constraints to enforce the generator network recover semantic texture details, we construct an end-to-end network to generate both the image contents and high-frequency details progressively. Our model contains a two-phase generator and two discriminators. In our generator, the previous phase restores the structure information via convolutional encoder-decoder architecture, and the following phase captures high-frequency details via residual learning. For each generator phase, we define different objective functions and further optimize the entire network via a feed-forward manner. Moreover, for the generator in the second phase, we adopt a deep residual architecture, which also can eliminate the perceptual discontinuity on the border of the missing region. Experimental results on several public datasets demonstrate qualitatively and quantitatively that our model performs better than the state-of-the-art algorithms and can generate both realistic image contents and high-frequency details. Our code will be released soon.

关键词： Mechatronics Machine vision Semantics Superresolution Stacking Noise reduction Linear programming