CVPR2020--几篇人群计数论文

Dear 丶 2023-02-21 06:22 1阅读 0赞

这里统计了CVPR2020的文章：[https://bbs.cvmart.net/topics/2751?from=groupmessage][https_bbs.cvmart.net_topics_2751_from_groupmessage]

## 1.CVPR2020\_Attention Scaling for Crowd Counting ##

作者 | Xiaoheng Jiang, Li Zhang, Mingliang Xu, Tianzhu Zhang, Pei Lv, Bing Zhou, Xin Yang, Yanwei Pang

单位 | 郑州大学；中国科学技术大学；大连理工大学；天津大学

摘要阅读笔记：

作者采用的网络结构：DANet、ASNet。

DANet为ASNet提供不同密度水平的感兴趣掩码（Attention mask），ASNet首先生成密度图和比例因子，然后将它们与感兴趣掩码相乘，以输出单独的基于注意的密度图。将这些密度图相加，得出最终的密度图。

作者提出了APLoss分层计算子区域的估计损失。

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzM1OTc1NDQ3_size_16_color_FFFFFF_t_70][]

## 2.CVPR2020\_Reverse Perspective Network for Perspective-Aware Object Counting ##

作者 | Yifan Yang, Guorong Li, Zhe Wu, Li Su, Qingming Huang, Nicu Sebe

单位 | UCAS；意大利特伦托大学等

摘要阅读笔记：

作者提出了反向透视网络（a reverse perspective network）解决一张图中尺度变化剧烈问题。

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzM1OTc1NDQ3_size_16_color_FFFFFF_t_70 1][]

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzM1OTc1NDQ3_size_16_color_FFFFFF_t_70 2][]

## 3.CVPR2020\_Adaptive Dilated Network With Self-Correction Supervision for Counting ##

作者 | Shuai Bai, Zhiqun He, Yu Qiao, Hanzhe Hu, Wei Wu, Junjie Yan

单位 | 北邮；商汤；中国科学院深圳先进技术研究院；北大

作者提出自纠正监督与自适应空洞网络，分别致力于改善类别标注不准确和尺度变化大的问题。

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzM1OTc1NDQ3_size_16_color_FFFFFF_t_70 3][]

## 参考 ##

[1.CVPR 2020 论文大盘点-行人检测与重识别篇][https_bbs.cvmart.net_topics_2751_from_groupmessage]

2.CVPR2020\_Attention Scaling for Crowd Counting

3.CVPR2020\_Reverse Perspective Network for Perspective-Aware Object Counting

4.CVPR2020\_Adaptive Dilated Network With Self-Correction Supervision for Counting

[5.CVPR 2020 open access（论文与代码下载）][5.CVPR 2020 open access]

[https_bbs.cvmart.net_topics_2751_from_groupmessage]: https://bbs.cvmart.net/topics/2751?from=groupmessage
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzM1OTc1NDQ3_size_16_color_FFFFFF_t_70]: https://img-blog.csdnimg.cn/20200629100238282.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzM1OTc1NDQ3,size_16,color_FFFFFF,t_70
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzM1OTc1NDQ3_size_16_color_FFFFFF_t_70 1]: https://img-blog.csdnimg.cn/20200629101554550.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzM1OTc1NDQ3,size_16,color_FFFFFF,t_70
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzM1OTc1NDQ3_size_16_color_FFFFFF_t_70 2]: https://img-blog.csdnimg.cn/20200629101348227.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzM1OTc1NDQ3,size_16,color_FFFFFF,t_70
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzM1OTc1NDQ3_size_16_color_FFFFFF_t_70 3]: https://img-blog.csdnimg.cn/20200629103355667.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzM1OTc1NDQ3,size_16,color_FFFFFF,t_70
[5.CVPR 2020 open access]: https://openaccess.thecvf.com/CVPR2020