In this repository, I only choose those crowd counting datasets used by top conferences (CVPR, ICCV and ECCV) papers, ordering by the release year of the datasets.
If you have any problems, suggestions or improvements, please submit the issue or PR.
In these papers published on top conferences(CVPR, ECCV) of 2020, some crowd counting benchmarks(datasets) are cited. In the following table, total reference count and release year of these datasets are showed.
| UCSD | UCF-CC-50 | WorldExpo10 | TRANCOS | ShangHai Tech | UCF-QNRF | |
|---|---|---|---|---|---|---|
| Ref Count 2020 | 2 | 7 | 7 | 2 | 10 | 7 |
| Release Year | 2008 | 2013 | 2015 | 2015 | 2016 | 2018 |
Note, only record the crowd datasets whose refered count is no less than 2 times.
Summary of statistics and comparason of the 4 most frequently used crowd datasets is presented in the following table.
| Dataset | Number of Images | Number of Annotations | Average Count | Maximum Count | Average Resolution | Average Density |
|---|---|---|---|---|---|---|
| UCF-CC-50 | 50 | 63,974 | 1279 | 4633 | 2101 x 2888 | 2.02 x 10^-4 |
| WorldExpo10 | 3980 | 225,216 | 56 | 334 | 576 x 720 | 1.36 x 10^-4 |
| ShanghaiTech_PartA | 482 | 241,677 | 501 | 3139 | 589 x 868 | 9.33 x 10^-4 |
| UCF-QNRF | 1535 | 1,251,642 | 815 | 12865 | 2013 x 2902 | 1.12 x 10^-4 |
Note, this table is abstracted from UCF-QNRF dataset website
Top Conference Papers using this dataset
- Yifan Yang, Guorong Li et al. Reverse Perspective Network for Perspective-Aware Object Counting. CVPR, 2020.
- Yifan Yang, Guorong Li et al. Weakly-Supervised Crowd Counting Learns from Sorting rather than Locations. ECCV, 2020.
Organization: Statistical Visual Computing Lab, UC San Diego
Website: http://svcl.ucsd.edu/projects/peoplecnt/
Paper: Antoni B Chan, Zhangsheng John Liang, and Nuno Vasconcelos. Privacy preserving crowd monitoring: Counting people without people models or tracking. Computer Vision and Pattern Recognition, pages 1–7, 2008.
Top Conference Papers using this dataset
- Xiaoheng Jiang, Li Zhang et al. Attention Scaling for Crowd Counting. CVPR, 2020. [code]
- Shuai Bai, Zhiqun He et al. Adaptive Dilated Network with Self-Correction Supervision for Counting. CVPR, 2020.
- Yutao Hu, Xiaolong Jiang et al. NAS-Count: Counting-by-Density with Neural Architecture Search. ECCV, 2020.
- Xiyang Liu, Jie Yang et al. Adaptive Mixture Regression Network with Local Counting Map for Crowd Counting. ECCV, 2020. [code]
- Liang Liu, Hao Lu et al. Weighing Counts: Sequential Crowd Counting by Reinforcement Learning? ECCV, 2020. [code]
- Vishwanath A. Sindagi, Rajeev Yasarla et al. Learning to Count in the Crowd from Limited Labeled Data. ECCV, 2020.
- Zhen Zhao, Miaojing Shi et al. Active Crowd Counting with Limited Supervision. ECCV, 2020.
Organization: Center for Research in Computer Vision, UCF
Website: https://www.crcv.ucf.edu/data/ucf-cc-50/
Paper: Haroon Idrees, Imran Saleemi, Cody Seibert, and Mubarak Shah. Multi-source multi-scale counting in extremely dense crowd images. Computer Vision and Pattern Recognition, pages 2547–2554, 2013.
Top Conference Papers using this dataset
- Xiaoheng Jiang, Li Zhang et al. Attention Scaling for Crowd Counting. CVPR, 2020. [code]
- Yutao Hu, Xiaolong Jiang et al. NAS-Count: Counting-by-Density with Neural Architecture Search. ECCV, 2020.
- Yifan Yang, Guorong Li et al. Reverse Perspective Network for Perspective-Aware Object Counting. CVPR, 2020.
- Xiyang Liu, Jie Yang et al. Adaptive Mixture Regression Network with Local Counting Map for Crowd Counting. ECCV, 2020. [code]
- Vishwanath A. Sindagi, Rajeev Yasarla et al. Learning to Count in the Crowd from Limited Labeled Data. ECCV, 2020.
- Yan Liu, Lingqiao Liu et al. Semi-Supervised Crowd Counting via Self-Training on Surrogate Tasks. ECCV, 2020.
- Yifan Yang, Guorong Li et al. Weakly-Supervised Crowd Counting Learns from Sorting rather than Locations. ECCV, 2020.
Organization: Institute of Image Communication and Network Engineering, Shanghai Jiao Tong University
Website: https://www.ee.cuhk.edu.hk/~xgwang/expo.html
Paper: Cong Zhang, Hongsheng Li, Xiaogang Wang, and Xiaokang Yang. Cross-scene crowd counting via deep convolutional neural networks. Computer Vision and Pattern Recognition, pages 833–841, 2015.
Top Conference Papers using this dataset
- Shuai Bai, Zhiqun He et al. Adaptive Dilated Network with Self-Correction Supervision for Counting. CVPR, 2020
- Zhen Zhao, Miaojing Shi et al. Active Crowd Counting with Limited Supervision. ECCV, 2020
Organization: GRAM, University of Alcala, Spain
Website: http://agamenon.tsc.uah.es/Personales/rlopez/data/trancos
Paper: Ricardo Guerrero-Gomez-Olmedo, Beatriz Torre-Jim ´enez, ´Roberto Lopez-Sastre, Saturnino Maldonado-Basc ´ on, and ´Daniel Onoro-Rubio. Extremely overlapping vehicle counting. In Iberian Conference on Pattern Recognition and Image Analysis, pages 423–431. Springer, 2015.
Top Conference Papers using this dataset
- Xiaoheng Jiang, Li Zhang et al. Attention Scaling for Crowd Counting. CVPR, 2020. [code]
- Shuai Bai, Zhiqun He et al. Adaptive Dilated Network with Self-Correction Supervision for Counting. CVPR, 2020.
- Yifan Yang, Guorong Li et al. Reverse Perspective Network for Perspective-Aware Object Counting. CVPR, 2020.
- Yutao Hu, Xiaolong Jiang et al. NAS-Count: Counting-by-Density with Neural Architecture Search. ECCV, 2020
- Xiyang Liu, Jie Yang et al. Adaptive Mixture Regression Network with Local Counting Map for Crowd Counting. ECCV, 2020. [code]
- Liang Liu, Hao Lu et al. Weighing Counts: Sequential Crowd Counting by Reinforcement Learning? ECCV, 2020. [code]
- Vishwanath A. Sindagi, Rajeev Yasarla et al. Learning to Count in the Crowd from Limited Labeled Data. ECCV, 2020.
- Yan Liu, Lingqiao Liu et al. Semi-Supervised Crowd Counting via Self-Training on Surrogate Tasks. ECCV, 2020.
- Zhen Zhao, Miaojing Shi et al. Active Crowd Counting with Limited Supervision. ECCV, 2020.
- Yifan Yang, Guorong Li et al. Weakly-Supervised Crowd Counting Learns from Sorting rather than Locations. ECCV, 2020.
Organization: ShanghaiTech University
Website: https://github.com/desenzhou/ShanghaiTechDataset
Paper: Yingying Zhang, Desen Zhou, Siqin Chen, Shenghua Gao, and Yi Ma. Single-image crowd counting via multi-column convolutional neural network. Computer Vision and Pattern Recognition, pages 589–597, 2016.
Top Conference Papers using this dataset
- Xiaoheng Jiang, Li Zhang et al. Attention Scaling for Crowd Counting. CVPR, 2020. [code]
- Shuai Bai, Zhiqun He et al. Adaptive Dilated Network with Self-Correction Supervision for Counting. CVPR, 2020.
- Yutao Hu, Xiaolong Jiang et al. NAS-Count: Counting-by-Density with Neural Architecture Search. ECCV, 2020.
- Xiyang Liu, Jie Yang et al. Adaptive Mixture Regression Network with Local Counting Map for Crowd Counting. ECCV, 2020. [code]
- Liang Liu, Hao Lu et al. Weighing Counts: Sequential Crowd Counting by Reinforcement Learning? ECCV, 2020. [code]
- Vishwanath A. Sindagi, Rajeev Yasarla et al. Learning to Count in the Crowd from Limited Labeled Data. ECCV, 2020.
- Yan Liu, Lingqiao Liu et al. Semi-Supervised Crowd Counting via Self-Training on Surrogate Tasks. ECCV, 2020.
Organization: Center for Research in Computer Vision, UCF
Website: https://www.crcv.ucf.edu/data/ucf-qnrf/
Paper: H. Idrees, M. Tayyab, K. Athrey, D. Zhang, S. Al-Maddeed, N. Rajpoot, M. Shah, Composition Loss for Counting, Density Map Estimation and Localization in Dense Crowds, in Proceedings of IEEE European Conference on Computer Vision (ECCV 2018), Munich, Germany, September 8-14, 2018.