成果/Result
- TSD-YOLO: Small traffic sign detection based on improved YOLO v8被引量：34收藏
- 作者：Du, Songjie Pan, Weiguo Li, Nuoya Dai, Songyin Xu, Bingxin Liu, Hongzhe Xu, Cheng Li, Xuewei
- 机构：Beijing Union Univ;Beijing Union Univ
- 来源：IET IMAGE PROCESSING 2024
- 关键词：computer vision convolutional neural nets object detection
- Bayesian Pseudoinverse Learners: From Uncertainty to Deterministic Learning被引量：7收藏
- 作者：Yin, Qian Xu, Bingxin Zhou, Kaiyan Guo, Ping
- 机构：Beijing Normal Univ;Beijing Union Univ;Beijing Normal Univ
- 来源：IEEE TRANSACTIONS ON CYBERNETICS 2022
- 关键词：Uncertainty Modeling Computational modeling Probabilistic logic Neural networks Bayes methods Task analysis Bayesian feedforward neural networks mixture priors pseudoinverse learners (PILs) synergetic learning
- PSC diffusion: patch-based simplified conditional diffusion model for low-light image enhancement被引量：7收藏
- 作者：Wan, Fei Xu, Bingxin Pan, Weiguo Liu, Hongzhe
- 机构：Beijing Union Univ;Beijing Union Univ
- 来源：MULTIMEDIA SYSTEMS 2024
- 关键词：Low-light image enhancement Generative model Diffusion model U-Net Image patching Parameter-free attention
- SFDiff: Diffusion model with sufficient spatial-Fourier frequency information interaction for low-light image enhancement被引量：6收藏
- 作者：Wan, Fei Xu, Bingxin Yao, Jingli Zeng, Lu Pan, Weiguo Liu, Hongzhe
- 机构：Beijing Union Univ;Beijing Union Univ;97 Beisihuan East Rd
- 来源：IET IMAGE PROCESSING 2024
- 关键词：diffusion fast Fourier transforms image enhancement image processing
- “学科思政”:探索研究生思政新模式被引量：6收藏
- 作者：刘宏哲 代松银 徐成 徐冰心 潘卫国
- 机构：北京联合大学;北京联合大学机器人学院
- 来源：《北京联合大学学报》 2020
- 关键词：学科思政 三全育人 思政合力 互助协作 资源融合
- 摘要：研究生思想政治教育作为高等学校思政教育的分支,在高校思政工作中占据重要的地位。通过对当前研究生思政工作中存在的问题进行分析,提出了以学科为平台开展思政工作,探索了各思政实施人员互助协作,课程、实验室和科研项目等资源融合,...
- Human Pose Estimation Based on Lightweight Multi-Scale Coordinate Attention被引量：5收藏
- 作者：Li, Xin Guo, Yuxin Pan, Weiguo Liu, Hongzhe Xu, Bingxin
- 机构：Beijing Union Univ
- 来源：APPLIED SCIENCES-BASEL 2023
- 关键词：human pose estimation attention mechanism multi-scale feature extraction
- TSIC-CLIP: Traffic Scene Image Captioning Model Based on Clip被引量：5收藏
- 作者：Zhang, Hao Xu, Cheng Xu, Bingxin Jiane, Muwei Liu, Hongzhe Li, Xuewei
- 机构：Beijing Union Univ;Beijing Union Univ;Shandong Univ Finance & Econ
- 来源：INFORMATION TECHNOLOGY AND CONTROL 2024
- 关键词：Contrastive learning Deep learning Image captioning Traffic scene Transformer
- Multi-channel expected patch log likelihood for color image denoising被引量：5收藏
- 作者：Zhou, Xiuling Xu, Bingxin Guo, Ping He, Ning
- 机构：Beijing City Univ;Beijing Union Univ;Beijing Normal Univ
- 来源：NEUROCOMPUTING 2019
- 关键词：Expected patch log-likelihood Multi-channel Matrix normal distribution Color image denoising
- 融合多尺度和注意力机制的小样本目标检测被引量：3收藏
- 作者：李鸿天 史鑫昊 潘卫国 徐成 徐冰心 袁家政
- 机构：北京市信息服务工程重点实验室(北京联合大学);北京开放大学科技学院
- 来源：《计算机应用》 2024
- 关键词：迁移学习 小样本目标检测 注意力机制 多尺度特征融合 余弦相似度
- 摘要：现有基于微调的二阶段小样本目标检测方法对新类特征不敏感,易将新类别误判成与它相似度高的基类,影响模型的检测性能。针对上述问题,提出一种融合多尺度和注意力机制的小样本目标检测(MA-FSOD)算法。首先在骨干网络使用分组卷...
- FSKT-GE: Feature maps similarity knowledge transfer for low-resolution gaze estimation被引量：3收藏
- 作者：Yan, Chao Pan, Weiguo Dai, Songyin Xu, Bingxin Xu, Cheng Liu, Hongzhe Li, Xuewei
- 机构：Beijing Union Univ;Beijing Union Univ
- 来源：IET IMAGE PROCESSING 2024
- 关键词：computer graphics computer vision convolutional neural nets
- Image Dehazing Based on Improved Color Channel Transfer and Multiexposure Fusion被引量：2收藏
- 作者：Ma, Shaojin Pan, Weiguo Liu, Hongzhe Dai, Songyin Xu, Bingxin Xu, Cheng Li, Xuewei Guan, Huaiguang
- 机构：Beijing Union Univ;Beijing Jiaotong Univ;CATARC Tianjin Automot Engn Res Inst Co Ltd
- 来源：ADVANCES IN MULTIMEDIA 2023
- 关键词：Color - Color image processing - Demulsification - Image enhancement - K-means clustering
- DAN: Distortion-aware Network for fisheye image rectification using graph reasoning被引量：2收藏
- 作者：Yan, Yongjia Liu, Hongzhe Zhang, Cheng Xu, Cheng Xu, Bingxin Pan, Weiguo Dai, Songyin Song, Yiqing
- 机构：Beijing Union Univ;Beijing Union Univ
- 来源：IMAGE AND VISION COMPUTING 2025
- 关键词：Fisheye image Distortion rectification Distortion pattern understanding Graph reasoning
- Deep Sparse Representation Classification with Stacked Autoencoder被引量：2收藏
- 作者：Xu, Bingxin Zhou, Xiuling
- 机构：Beijing Union Univ;Beijing City Univ
- 来源：2019 15TH INTERNATIONAL CONFERENCE ON COMPUTATIONAL INTELLIGENCE AND SECURITY (CIS 2019) 2019
- 关键词：deep sparse representation stacked autoencoder pseudoinverse learning dictionary Learning
- Wavelet-Based Texture Mining and Enhancement for Face Forgery Detection被引量：1收藏
- 作者：Li, Xin Zhao, Hui Xu, Bingxin Liu, Hongzhe
- 机构：Beijing Union Univ;Beijing Union Univ
- 来源：IET BIOMETRICS 2025
- 关键词：face forgery detection frequency-guided texture enhancement image frequency analysis multiscale feature interaction
- DBYOLOv8: Dual-Branch YOLOv8 Network for Small Object Detection on Drone Image被引量：1收藏
- 作者：Tan, Yawei Xu, Bingxin Sun, Jiangsheng Xu, Cheng Pan, Weiguo Dai, Songyin Liu, Hongzhe
- 机构：Beijing Union Univ;Army Res Acad
- 来源：INTERNATIONAL JOURNAL OF ADVANCED COMPUTER SCIENCE AND APPLICATIONS 2025
- 关键词：Drone images dual-branch small object detection YOLOv8
- Combined spatial and frequency dual stream network for face forgery detection被引量：1收藏
- 作者：Zhao, Hui Li, Xin Xu, Bingxin Liu, Hongzhe
- 机构：Beijing Union Univ;Beijing Union Univ
- 来源：PEERJ COMPUTER SCIENCE 2024
- 关键词：Face forgery detection Multi-scale feature extraction Cross self attention Image frequency analysis
- 面向复杂场景的细粒度图像分类研究进展综述被引量：0收藏
- 作者：汪金宇 曾凤彩 代松银 徐冰心
- 机构：北京联合大学北京市信息服务工程重点实验室;北京联合大学机器人学院脑与认知智能北京实验室
- 来源：《中国计算机用户协会网络应用分会2024年第二十八届网络新技术与应用年会》 2024
- 关键词：细粒度图像分类 深度学习 复杂场景 特征提取 多尺度特征学习
- 摘要：细粒度图像分类旨在区分同一大类中具有微小差异的子类对象,被广泛应用于车辆、植物、鸟类等领域。然而,复杂场景下的细粒度图像分类面临诸多挑战,如背景干扰、光照变化、遮挡、视角变化以及相似类别间的微小差异。近年来,随着深度学习...
- Low-Light Image Enhancement using Retinex-based Network with Attention Mechanism被引量：0收藏
- 作者：Ma, Shaojin Pan, Weiguo Li, Nuoya Du, Songjie Liu, Hongzhe Xu, Bingxin Xu, Cheng Li, Xuewei
- 机构： Beijing Key Laboratory of Information Service Engineering; College of Robotics
- 来源：International Journal of Advanced Computer Science and Applications 2024
- 关键词：Finite element method
- 文物壁画的智能数字化修复研究被引量：0收藏
- 作者：徐成 孙吉武 王石磊 徐冰心 刘宏哲
- 机构：北京联合大学北京市信息服务重点实验室;北京联合大学机器人学院脑与认知智能北京实验室;北京联合大学应用文理学院法律系
- 来源：《中国计算机用户协会网络应用分会2023年第二十七届网络...》 2023
- 关键词：文物壁画 文化遗产 数字化修复 深度学习 虚拟现实
- 摘要：文物壁画是历史文化的重要载体和传承者。我国拥有众多的文物壁画,这些壁画具有重要的历史和文化价值。壁画类文化遗产目前存在多种病害,例如起甲、龟裂、疱疹、酥碱和空臌等,并且易发生颜料层脱落、褪色等现象。为保护这些文化遗产,数...
- Pseudoinverse Learning Algorithom for Fast Sparse Autoencoder Training被引量：0收藏
- 作者：Xu, Bingxin Guo, Ping
- 机构：Beijing Union Univ;Beijing Normal Univ
- 来源：2018 IEEE CONGRESS ON EVOLUTIONARY COMPUTATION (CEC) 2018
- 关键词：sparse autoencoder pseudoinverse learning algorithm fast learning data representation
- Reconstructing damaged fNIRS signals with a generative deep learning model被引量：0收藏
- 作者：Zhi, Yingxu Zhang, Baiqiang Xu, Bingxin Wan, Fei Niu, Peisong Niu, Haijing
- 机构： State Key Laboratory of Cognitive Neuroscience and Learning & IDG/McGovern Institute for Brain Research; Beijing Key Laboratory of Information Service Engineering
- 来源：Artificial Intelligence Review 2025
- 关键词：Deep reinforcement learning - Generative adversarial networks - Signal to noise ratio
- Technique Report of CVPR 2024 PBDL Challenges被引量：0收藏
- 作者：Lu, Xiaoqiang Jiao, Licheng Liu, Fang Liu, Xu Li, Lingling Ma, Wenping Yang, Shuyuan Xie, Haiyang Zhao, Jian Huang, Shihuang Cheng, Peng Shen, Xi Wang, Zheng An, Shuai Zhu, Caizhi Li, Xuelong Chen, Linwei Fu, Ying Zhang, Tao Li, Liang Liu, Yu Yan, Chenggang Wang, Zichun Wang, Qinliang Gou, Xuejian Liu, Yang Yu, Xinyue Jia, Sen Zhang, Junpei Jiao, Licheng Liu, Xu Chen, Puhua Chen, Xiang Li, Hao Pan, Jinshan Xie, Chuanlong Chen, Hongming Li, Mingrui Deng, Tianchen Huang, Jingwei Li, Yufeng Wan, Fei Xu, Bingxin Cheng, Jian Liu, Hongzhe Xu, Cheng Zou, Yuxiang Pan, Weiguo Dai, Songyin Jiang, Linyan Song, Bingyi An, Zhuoyu Lei, Haibo Luo, Qing Song, Jie Liu, Yuan Li, Qihang Zhang, Haoyuan Wang, Lingfeng Chen, Wei Luo, Aling Li, Cheng Cao, Jun Chen, Shu Dou, Zifei Liu, Xinyu Zhang, Jing Zhang, Kexin Yang, Yuting Yang, Shuyuan Zhang, Liwen Xu, Zhe Gou, Dingyong Li, Cong Xu, Senyan Zhang, Yunkang Jiang, Siyuan Liu, Qinglin Yu, Wei Lv, Xiaoqian Li, Jianing Zhang, Shengping Ji, Xiangyang Zou, Yunhao Chen, Yuanpei Zhang, Yuhan Peng, Weihang Zhao, Shizhan Zhang, Yanzhao Yan, Libo Lu, Xiaoqiang Guo, Yuwei Li, Guoxin Gao, Qiong Che, Chenyue Sun, Long
- 机构： School of Artificial Intelligence; School of Computer Science; Intellindust; Beijing Forestry University; Institute of AI
- 来源：arXiv 2024
- 关键词：High energy physics
- Optimization of Exploration Prospects Based on Ant Colony Algorithm and XGBoost Combined Optimization Model被引量：0收藏
- 作者：Song, Mengxin Xu, Bingxin Feng, Mei Fu, Xinxi
- 机构： Research Institute of Petroleum Exploration and Development RIPED; Beijing Key Laboratory of Information Service Engineering
- 来源：Society of Petroleum Engineers - Abu Dhabi International Petroleum Exhibition and Conference, ADIP 2021 2021
- 关键词：Gasoline - Learning systems - Classification (of information) - Ant colony optimization - Decision support systems - Forecasting - Geology - Petroleum prospecting
- Broad and Pseudoinverse Learning for Autoencoder被引量：0收藏
- 作者：Xu, Bingxin Guo, Ping
- 机构：Beijing Union Univ;Beijing Normal Univ
- 来源：2018 IEEE INTERNATIONAL CONFERENCE ON SYSTEMS, MAN, AND CYBERNETICS (SMC) 2018
- 关键词：Autoencoder Pseudoinverse learning algorithm Receptive function Broad learning
- Self-attention enhanced dynamic semantic multi-scale graph convolutional network for skeleton-based action recognition被引量：0收藏
- 作者：Liu, Shihao Xu, Cheng Dai, Songyin Li, Nuoya Pan, Weiguo Xu, Bingxin Liu, Hongzhe
- 机构：Beijing Union Univ
- 来源：IMAGE AND VISION COMPUTING 2025
- 关键词：Skeleton-based action recognition Graph convolutional network Self-attention Multi-scale modeling
- 图像描述生成的研究进展综述被引量：0收藏
- 作者：张浩 贯怀光 徐成 徐冰心 潘卫国
- 机构：北京联合大学北京市信息服务工程重点实验室;北京联合大学机器人学院脑与认知智能北京实验室;中汽研(天津)汽车工程研究院有限公司
- 来源：《中国计算机用户协会网络应用分会2022年第二十六届网络...》 2022
- 关键词：图像描述生成 自然语言处理 深度学习 卷积神经网络 循环神经网络
- 摘要：图像描述属于深度学习中比较热门一个研究方向,是多角度信息对齐任务。通过对图像信息与文本信息建模,使得计算机能够识别图像的语义信息,并生成相应的人们可以理解的自然语言文本。图像描述作为计算机视觉与自然语言处理领域的任务,在...
- Generalization-oriented face forgery detection via discriminative feature analysis and normalization被引量：0收藏
- 作者：Li, Xin Xu, Bingxin Liu, Hongzhe Pan, Weiguo Xu, Cheng
- 机构： Beijing Key Laboratory of Information Service Engineering; College of Robotics
- 来源：Multimedia Systems 2025
- 关键词：Data mining - Feature extraction - Textures
- 基于深度学习的低光照图像增强方法研究综述被引量：0收藏
- 作者：李诺亚 潘卫国 刘世豪 徐冰心
- 机构：北京联合大学北京市信息服务工程重点实验室;北京联合大学机器人学院脑与认知智能北京实验室
- 来源：《中国计算机用户协会网络应用分会2024年第二十八届网络新技术与应用年会》 2024
- 关键词：深度学习 低光照图像增强 CNN Retinex GAN
- 摘要：低光照图像增强旨在提高光照不足场景下捕获数据的视觉感知质量以获得更多有效信息,其已成为图像处理领域的研究热点,在自动驾驶等人工智能产业中具有广阔的应用前景。传统方法的推导过程流程普遍复杂,不利于实际应用。当前,基于深度学...
- DCDFusion: A dual-consistency decoupling infrared and visible image fusion algorithm for low-light scenes被引量：0收藏
- 作者：Xie, Wenkuan Pan, Weiguo Zhang, Jiancheng Dai, Songyin Xu, Bingxin Xu, Cheng
- 机构：Beijing Union Univ
- 来源：INFRARED PHYSICS & TECHNOLOGY 2026
- 关键词：Low-light scenes Infrared and visible image fusion Low-light adaptive enhancement Dual-consistency Object detection
- 基于Faster RCNN的智能车道路前方车辆检测方法被引量：0收藏
- 作者：史凯静 鲍泓 徐冰心 潘卫国 郑颖
- 机构：北京联合大学北京市信息服务工程重点实验室
- 来源：《计算机工程》 2018
- 关键词：智能车 前方车辆 深度卷积神经网络 训练模型 准确率
- 摘要：使用Fast RCNN方法进行特征提取存在耗时较长且检测准确率较低的问题。为此,结合Faster RCNN前方车辆检测模型与3种不同大小的卷积神经网络,提出一种改进的前方车辆检测方法,研究对比各方法在3种交通场景数据库上...
