1,583 Hits in 5.2 sec

Learning Multi-level Features to Improve Crowd Counting

Zhanqiang Huo, Bin Lu, Aizhong Mi, Fen Luo, Yingxu Qiao
2020 IEEE Access  
CROWD COUNTING In recent years, crowd counting methods have made great progress by employing convolutional neural networks to regress crowd density maps.  ...  [15] proposed a Perspective Crowd Counting via Spatial Convolutional Network (PCC Net) to solve high appearance similarity, perspective changes and severe congestion.  ... 
doi:10.1109/access.2020.3039998 fatcat:tz4f6ws42fem3ldx727uutsuzm

VCIP 2020 Index

2020 2020 IEEE International Conference on Visual Communications and Image Processing (VCIP)  
Fusion Network for Anomaly Detection Wang, Yongfang Enhanced Saliency Prediction via Orientation Selectivity Wang, Yuchen Deep Convolutional Neural Network Based on Multi-Scale Feature Extraction  ...  Spatial-Channel Context-Based Entropy Modeling for End-to-end Optimized Image Compression Li, Chun-Guang Learning Convolution Feature Aggregation via Edge Attention Convolution Network for Perso  ... 
doi:10.1109/vcip49819.2020.9301896 fatcat:bdh7cuvstzgrbaztnahjdp5s5y

An Automatic Scale-adaptive Approach with Attention Mechanism-based Crowd Spatial Information for Crowd Counting

Weihang Kong, He Li, Guanglong Xing, Fengda Zhao
2019 IEEE Access  
the main crowd counting network.  ...  Then, the semantic feature of the classification network is converted into the crowd spatial information mask via the proposed spatial attention conversion module, and the crowd spatial information mask  ...  The initial work on crowd counting mainly adopts handcrafted features, and with the recent development of convolutional neural network (CNN), remarkable progress has occurred in the crowd counting tasks  ... 
doi:10.1109/access.2019.2918936 fatcat:mcrvyp6opvavvkaxyirh3pgqsu

A Survey on Deep Learning-based Single Image Crowd Counting: Network Design, Loss Function and Supervisory Signal [article]

Haoyue Bai, Jiageng Mao, S.-H. Gary Chan
2022 arXiv   pre-print
After presenting publicly available datasets and evaluation metrics, we review the recent advances with detailed comparisons on three major design modules for crowd counting: deep neural network designs  ...  This survey is to provide a comprehensive summary of recent advances on deep learning-based crowd counting techniques via density map estimation by systematically reviewing and summarizing more than 200  ...  . • Graph neural networks based method distills rich relations among multi-scale features for crowd counting.  ... 
arXiv:2012.15685v2 fatcat:kvrqnczkgbdxdnvr4243atvj3e

Enhanced 3D convolutional networks for crowd counting [article]

Zhikang Zou, Huiliang Shao, Xiaoye Qu, Wei Wei, Pan Zhou
2019 arXiv   pre-print
Recently, convolutional neural networks (CNNs) are the leading defacto method for crowd counting.  ...  By stacking TCA blocks together, we obtain the deep trainable architecture called enhanced 3D convolutional networks (E3D).  ...  Recently, researchers have leveraged Convolutional Neural Networks (CNNs) for an accurate crowd density map generation [2, 16, 33, 37] .  ... 
arXiv:1908.04121v1 fatcat:wpfdr3hy5fdi5gsp2lynkmaluy

Robust Crowd Counting via Image Enhancement and Dynamic Feature Selection

Nayeong Kim, Suha Kwak
2021 British Machine Vision Conference  
In spite of their remarkable success in many vision tasks, convolutional neural networks (CNNs) often has trouble counting people in crowded scenes due to the following reasons.  ...  First, we develop a new counting network called pyramid feature selection network (PFSNet) that adapts its receptive fields dynamically to local crowd densities of the input image.  ...  Recently, convolutional neural networks (CNNs) have been widely adopted for the direct prediction of the count.  ... 
dblp:conf/bmvc/KimK21 fatcat:zx2jv2pu7vhpbi5o6lrlsborkq

HAGN: Hierarchical Attention Guided Network for Crowd Counting

Zuodong Duan, Yujun Xie, Jiahao Deng
2020 IEEE Access  
One approach adopts different network structures to model scale variation, such as Multi-column Convolutional Neural Network (MCNN) [25] , Switching Convolutional Neural Network (Switch CNN) [26] and  ...  [38] proposed the Convolutional Block Attention Module (CBAM), which could achieve attentional calculation of spatial and channel dimensions in the forward process of convolutional neural networks.  ... 
doi:10.1109/access.2020.2975268 fatcat:6weivtgdw5bktpybqzmeknpbcm

SOFA-Net: Second-Order and First-order Attention Network for Crowd Counting [article]

Haoran Duan, Shidong Wang, Yu Guan
2020 arXiv   pre-print
To obtain the appropriate crowd representation, in this work we proposed SOFA-Net(Second-Order and First-order Attention Network): second-order statistics were extracted to retain selectivity of the channel-wise  ...  spatial information for dense heads while first-order statistics, which can enhance the feature discrimination for the heads' areas, were used as complementary information.  ...  It is the most widely used loss function on training deep convolution networks in crowd counting tasks [12, 14, 20] .  ... 
arXiv:2008.03723v1 fatcat:vcpufazxhrhjri7qpc5aiudvwy

Multi-scale Feature Adaptive Integration for Crowd Counting in Highly Congested Scenes

Hui Gao, Miaolei Deng, Wenjun Zhao, Dexian Zhang, Yuehong Gong
2022 IEEE Access  
In this paper, we propose a new multi-scale feature adaptive integrated network (MSFAINet) for crowd counting that adopts the multiscale feature, hybrid attention, and dilated convolution.  ...  Second, it adopts a hybrid attention mechanism to enhance the receptive field of an image while reducing the loss of feature information caused by channel competition and then passes these features into  ...  [27] proposed a scale-driven convolutional neural network (SD-CNN), which is based on the assumption that heads are the dominant and visible features regardless of the density of crowds.  ... 
doi:10.1109/access.2022.3171328 fatcat:rqh7zuekbvd35hrevojkpwy2ce

Multi‐level features extraction network with gating mechanism for crowd counting

Xin Zeng, Qiang Guo, Haoran Duan, Yunpeng Wu
2021 IET Image Processing  
Unlike previous works, a multi-level features extraction network with gating mechanism for crowd counting is proposed.  ...  Most existing methods based on the straightforward fusion of different features from a deep neural network seem to eliminate this limitation.  ...  Compared to traditional approaches, convolutional neural network (CNN) has powerful feature representation ability.  ... 
doi:10.1049/ipr2.12304 fatcat:gxx5eesitrasfeuympkqrrmwfq

Crowd counting via Multi-Scale Adversarial Convolutional Neural Networks

Liping Zhu, Hong Zhang, Sikandar Ali, Baoli Yang, Chengyang Li
2020 Journal of Intelligent Systems  
In this paper, a novel end-to-end architecture based on Multi-Scale Adversarial Convolutional Neural Network (MSA-CNN) is proposed to generate crowd density and estimate the amount of crowd.  ...  Current methods solve these issues by compounding multi-scale Convolutional Neural Network with different receptive fields.  ...  To solve these issues based on the multi-column CNN [19] which has a success of working in the crowd counting, a new crowd counting framework called Multi-Scale Adversarial Convolutional Neural Network  ... 
doi:10.1515/jisys-2019-0157 fatcat:bhj42iinorekdcz2jruryus5uq

Convolutional Neural Network for Crowd Counting on Metro Platforms

Jun Zhang, Jiaze Liu, Zhizhong Wang
2021 Symmetry  
In this paper, in order to solve the problem of metro platform passenger flow detection, we propose a CNN (convolutional neural network)-based network called the MP (metro platform)-CNN to accurately count  ...  The proposed method is composed of three major components: a group of convolutional neural networks is used on the front end to extract image features, a multiscale feature extraction module is used to  ...  Therefore, convolutional-neural-network-based crowd counting methods are more suitable for use on metro platforms.  ... 
doi:10.3390/sym13040703 fatcat:w2d7p3igdnbjdicijeo4clnyfy

Crowd Counting via Hierarchical Scale Recalibration Network [article]

Zhikang Zou and Yifan Liu and Shuangjie Xu and Wei Wei and Shiping Wen and Pan Zhou
2020 arXiv   pre-print
The task of crowd counting is extremely challenging due to complicated difficulties, especially the huge variation in vision scale.  ...  In order to reallocate channel-wise feature responses, a Scale Recalibration Module (SRM) adopts a step-by-step fusion to generate final density maps.  ...  With the rapid growth of convolutional neural networks, many CNN-based methods [18, 1, 25] have sprung up in fields of crowd counting and have made promising progress.  ... 
arXiv:2003.03545v1 fatcat:23vtojoeezao7gtbhxfj2xwb5i

Convolutional-Neural Network-Based Image Crowd Counting: Review, Categorization, Analysis, and Performance Evaluation

Naveed Ilyas, Ahsan Shahzad, Kiseon Kim
2019 Sensors  
We also highlight the potential applications of convolutional-neural-network-based crowd-counting techniques.  ...  In this article, we review, categorize, analyze (limitations and distinctive features), and provide a detailed performance evaluation of the latest convolutional-neural-network-based crowd-counting techniques  ...  The authors in [26] focused on conventional and convolutional-neural-network (CNN)-based single-image crowd-counting techniques.  ... 
doi:10.3390/s20010043 pmid:31861734 pmcid:PMC6983207 fatcat:gvso42grpjbw5ptdb23sdfeuwu

SCFFNet: Spatial Context Feature Fusion Network for Understanding the Highly Congested Scenes

Liyan Xiong, Hu Yi, Xiaohui Huang, Weichun Huang, Nouman Ali
2022 Mathematical Problems in Engineering  
To solve these problems, this paper proposes a spatial context feature fusion network, abbreviated as SCFFNet, to understand highly congested scenes and perform accurate counts as well as produce high-quality  ...  calibrates and refuses the fused feature maps through a channel spatial attention-aware module, which improves the model's ability to suppress background and focus on main features.  ...  CSRNet uses the deep convolutional neural network VGG-16 net [14] that removes the fully connected layer as the feature extractor, followed by a 7-layer dilated convolution as the regression, which can  ... 
doi:10.1155/2022/3277995 fatcat:42tnlg5hebgpjh4k3mfqh7l3cq
« Previous Showing results 1 — 15 out of 1,583 results