Crowd Counting with Deep Negative Correlation Learning

Shi, Zenglin; Zhang, Le; Liu, Yun; Cao, Xiaofeng; Ye, Yangdong; Cheng, Ming–Ming; Zheng, Guoyan

doi:10.1109/cvpr.2018.00564

Cited by 274 publications

(164 citation statements)

References 28 publications

Supporting

Mentioning

161

Contrasting

Order By: Relevance

“…In the past decade, a number of crowd counting algorithms [22,58,21,12,5,35,20,11,7,31] have been proposed in the literature. Recently, crowd counting methods using Convolutional Neural Networks (CNNs) have made remarkable progresses [53,46,36,6,57,59,38,9,4,32,16]. The best performing methods are mostly based on the density map estimation, which typically obtain the crowd count by predicting a density map for the input image and then summing over the estimated density map.…”

Section: Introductionmentioning

confidence: 99%

Bayesian Loss for Crowd Count Estimation With Point Supervision

Wei

Hong

et al. 2019

2019 IEEE/CVF International Conference on Computer Vision (ICCV)

475

344

View full text Add to dashboard Cite

In crowd counting datasets, each person is annotated by a point, which is usually the center of the head. And the task is to estimate the total count in a crowd scene. Most of the state-of-the-art methods are based on density map estimation, which convert the sparse point annotations into a "ground truth" density map through a Gaussian kernel, and then use it as the learning target to train a density map estimator. However, such a "ground-truth" density map is imperfect due to occlusions, perspective effects, variations in object shapes, etc. On the contrary, we propose Bayesian loss, a novel loss function which constructs a density contribution probability model from the point annotations. Instead of constraining the value at every pixel in the density map, the proposed training loss adopts a more reliable supervision on the count expectation at each annotated point. Without bells and whistles, the loss function makes substantial improvements over the baseline loss on all tested datasets. Moreover, our proposed loss function equipped with a standard backbone network, without using any external detectors or multi-scale architectures, plays favourably against the state of the arts. Our method outperforms previous best approaches by a large margin on the latest and largest UCF-QNRF dataset. The source code is available at https://github.com/ZhihengCV/ Baysian-Crowd-Counting. arXiv:1908.03684v1 [cs.CV] 10 Aug 2019 pixel, our proposed training loss supervises on the count expectation at each annotated point, instead.Extensive experimental evaluations show that the proposed loss function substantially outperforms the baseline training loss on UCF-QNRF [16], ShanghaiTech [57], and UCF CC 50 [15] benchmark datasets. Moreover, our proposed loss function equipped with the standard VGG-19 network [39] as backbone, without using any external detectors or multi-scale architectures, achieves the state-ofthe-art performances on all the benchmark datasets, especially with a magnificent improvement on the UCF-QNRF dataset compared to other methods.

show abstract

Section: Introductionmentioning

confidence: 99%

Bayesian Loss for Crowd Count Estimation With Point Supervision

Wei

Hong

et al. 2019

2019 IEEE/CVF International Conference on Computer Vision (ICCV)

475

344

View full text Add to dashboard Cite

show abstract

“…The proposed method also performs slightly better than L2R [18] in transferring mod-els trained on ShanghaiTech Part A to UCF CC 50. Yet, the improvement is not as significant as the comparison with [33,25] on transferring between ShanghaiTech Part A and Part B. This is probably because L2R [18] also relies on extra data which may somehow help to reduce the gap between the two datasets.…”

Section: Evaluation Of Transferabilitymentioning

confidence: 91%

“…To overcome the over-fitting, Liu et al [18] propose a learning-torank framework to leverage abundantly available unlabeled crowd images and a self-learning strategy. Shi et al [25] build a set of decorrelated regressors with reasonable generalization capabilities through managing their intrinsic diversities to avoid severe over-fitting. Though many methods have been proposed to tackle the large scale and density variation issue, this problem still remains challenging for crowd counting.…”

Section: Related Workmentioning

confidence: 99%

Learn to Scale: Generating Multipolar Normalized Density Maps for Crowd Counting

Qiu

et al. 2019

2019 IEEE/CVF International Conference on Computer Vision (ICCV)

140

View full text Add to dashboard Cite

Dense crowd counting aims to predict thousands of human instances from an image, by calculating integrals of a density map over image pixels. Existing approaches mainly suffer from the extreme density variances. Such density pattern shift poses challenges even for multi-scale model ensembling. In this paper, we propose a simple yet effective approach to tackle this problem. First, a patch-level density map is extracted by a density estimation model and further grouped into several density levels which are determined over full datasets. Second, each patch density map is automatically normalized by an online center learning strategy with a multipolar center loss. Such a design can significantly condense the density distribution into several clusters, and enable that the density variance can be learned by a single model. Extensive experiments demonstrate the superiority of the proposed method. Our work outperforms the state-of-the-art by 4.Sparse Medium Dense (a) Shanghai-B Shanghai-A UCF_CC_50 UCF-QNRF Dataset 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 Mean Relative Error MRE Overview Ours MCNN [33] Switch-CNN [23] CMTL [27] ACSCP [24] SANet [3] CSRNet [15] CP-CNN [28] L2R [18] D-ConvNet-v1 [25] ic-CNN [22] arXiv:1907.12428v2 [cs.CV]

show abstract

“…Due to these issues, crowd counting and density estimation is a very difficult problem, especially in highly congested scenes. Several recent convolutional neural network (CNN) based methods for counting [1,2,3,4,5,6,7,8,9,10,11,12] have attempted to address one or more of these issues by adding more robustness to scale variations by proposing different techniques such as multicolumn networks [2], intelligent selection of regressors suited for a particular crowd scenario [3] and incorporating global, local context information into the counting network [4], etc. Methods such as [3,4,8] achieve significantly lower errors compared to the earlier approaches, however, they are complex to train due to the presence of multiple learning stages.…”

Section: Introductionmentioning

confidence: 99%

“…Similarly, CP-CNN [4] requires that their local and global estimators to be trained separately, followed by end-toend training of their density map estimator. Although the most recent methods such as [6,7,10] achieve better results while being efficient, there is still considerable room for further improvements. In this paper, we propose to improve the counting performance by explicitly modeling spatial pixel-wise attention and global attention into the counting network.…”

Section: Introductionmentioning

confidence: 99%

HA-CCN: Hierarchical Attention-Based Crowd Counting Network

Sindagi

Patel

2020

IEEE Trans. on Image Process.

184

View full text Add to dashboard Cite

Single image-based crowd counting has recently witnessed increased focus, but many leading methods are far from optimal, especially in highly congested scenes. In this paper, we present Hierarchical Attention-based Crowd Counting Network (HA-CCN) that employs attention mechanisms at various levels to selectively enhance the features of the network. The proposed method, which is based on the VGG16 network, consists of a spatial attention module (SAM) and a set of global attention modules (GAM). SAM enhances low-level features in the network by infusing spatial segmentation information, whereas the GAM focuses on enhancing channel-wise information in the higher level layers. The proposed method is a single-step training framework, simple to implement and achieves state-of-the-art results on different datasets. Furthermore, we extend the proposed counting network by introducing a novel set-up to adapt the network to different scenes and datasets via weak supervision using image-level labels. This new set up reduces the burden of acquiring labour intensive point-wise annotations for new datasets while improving the cross-dataset performance.

show abstract

Crowd Counting with Deep Negative Correlation Learning

Cited by 274 publications

References 28 publications

Bayesian Loss for Crowd Count Estimation With Point Supervision

Bayesian Loss for Crowd Count Estimation With Point Supervision

Learn to Scale: Generating Multipolar Normalized Density Maps for Crowd Counting

HA-CCN: Hierarchical Attention-Based Crowd Counting Network

Contact Info

Product

Resources

About