Awesome Crowd Localization
- Crowd Counting
- Crowd Analysis
- Video Surveillance
- Dense/Small/Tiny Object Detection
- mAP, mAR in RAZNet (namely key point evaluation in COCO: fixed sigma)
- F1-m, Precision, Recall in NWPU-Crowd (scale-aware sigma)
- MLE in LSC-CNN (distance measure)
- NWPU-Crowd (dot, box)
- JHU-CROWD (dot, size)
- FDST (dot, box)
- Head Tracking 21 (dot, box, id) [Download]
- [DCST] Congested Crowd Instance Localization with Dilated Convolutional Swin Transformer [paper]
- [GNA] Video Crowd Localization with Multi-focus Gaussian Neighbor Attention and a Large-Scale Benchmark [paper]
- [SCALNet] Dense Point Prediction: A Simple Baseline for Crowd Counting and Localization [paper] [code]
- [FIDTM] Focal Inverse Distance Transform Maps for Crowd Localization and Counting in Dense Crowd [paper] [code]
- [RDTM] Reciprocal Distance Transform Maps for Crowd Counting and People Localization in Dense Crowd [paper] [code]
- Counting and Locating High-Density Objects Using Convolutional Neural Network [paper]
- [IIM] Learning Independent Instance Maps for Crowd Localization [paper] [code]
- [AutoScale] Autoscale: learning to scale for crowd counting [paper] [code]
- A Strong Baseline for Crowd Counting and Unsupervised People Localization [paper]
- Drone-based Joint Density Map Estimation, Localization and Tracking with Space-Time Multi-Scale Attention Network [paper][code]
- [SA-InterNet] A-InterNet: Scale-Aware Interaction Network for Joint Crowd Counting and Localization (PRVC) [paper]
- A smartly simple way for joint crowd counting and localization (Neurocomputing) [aper]
- A Generalized Loss Function for Crowd Counting and Localization (CVPR) [paper]
- [ P2PNet] Rethinking Counting and Localization in Crowds: A Purely Point-Based Framework (ICCV) [paper]
- [D2CNet] Decoupled Two-Stage Crowd Counting and Beyond (TIP) [paper] [code]
- [Crowd-SDNet] A Self-Training Approach for Point-Supervised Object Detection and Counting in Crowds (TIP) [paper] [code]
- [TopoCount] Localization in the Crowd with Topological Constraints (AAAI2021) [paper][code]
- [DD-CNN] Going Beyond the Regression Paradigm with Accurate Dot Prediction for Dense Crowds (WACV) [paper]
- [NWPU] NWPU-Crowd: A Large-Scale Benchmark for Crowd Counting and Localization (T-PAMI) [paper][code]
- [LSC-CNN] Locate, Size and Count: Accurately Resolving People in Dense Crowds via Detection (T-PAMI) [paper][code]
- Scale Match for Tiny Person Detection (WACV) [paper][code]
- Density Map Regression Guided Detection Network for RGB-D Crowd Counting and Localization (CVPR) [paper]
- Point in, Box out: Beyond Counting Persons in Crowds (CVPR) [paper]
- [RAZ_Loc] Recurrent attentive zooming for joint crowd counting and precise localization (CVPR) [paper] [Reproduction_code]
- [RDNet] Density Map Regression Guided Detection Network for RGB-D Crowd Counting and Localization (CVPR) [paper][code]
- [CL] Composition Loss for Counting, Density Map Estimation and Localization in Dense Crowds (ECCV) [paper]
- [LCFCN] Where are the Blobs: Counting by Localization with Point Supervision (ECCV) [paper] [code]
- SOD-MTGAN: Small Object Detection via Multi-Task Generative Adversarial Network (ECCV) [paper]
- Focal Loss for Dense Object Detection (ICCV) [paper]
- [TinyFaces] Finding tiny faces (CVPR) [paper]
- Perceptual Generative Adversarial Networks for Small Object Detection (CVPR) [paper]
- Small Instance Detection by Integer Programming on Object Density Maps, (CVPR) [paper ]
- End-to-end people detection in crowded scenes (CVPR) [paper] [code]
- [Faster-RCNN] Towards real-time object detection with region proposal networks (CVPR) [paper] [code]
More detailed results are in this link.
| Year--Conference/Journal | Methods | Backbone | F1-measure | Precise | Recall | A0~A5 | Avg. |
|---|---|---|---|---|---|---|---|
| 2015--NIPS | Faster RCNN | ResNet-101 | 6.7 | 95.8 | 3.5 | 0/0.002/0.4/7.9/37.2/63.5 | 18.2 |
| 2017--CVPR | TinyFaces | ResNet-101 | 56.7 | 52.9 | 61.1 | 4.2/22.6/59.1/90.0/93.1/89.6 | 59.8 |
| 2019--arXiv | VGG+GPR | VGG-16 | 52.5 | 55.8 | 49.6 | 3.1/27.2/49.1/68.7/49.8/26.3 | 37.4 |
| 2019--CVPR | RAZ_Loc | VGG-16 | 59.8 | 66.6 | 54.3 | 3.1/27.2/49.1/68.7/49.8/26.3 | 42.4 |
| 2021--TIP | Crowd-SDNet | ResNet-50 | 63.7 | 65.1 | 62.4 | 7.3/43.7/62.4/75.7/71.2/70.2 | 55.1 |
| 2021--AAAI | TopoCount | VGG-16 | 69.2 | 68.3 | 70.1 | 5.7/39.1/72.2/85.7/87.3/89.7 | 63.3 |
| 2021--arXiv | RDTM | VGG-16 | 69.9 | 75.1 | 65.4 | 11.5/46.3/68.5/74.9/54.6/18.2 | 45.7 |
| 2021--arXiv | SCALNet | DLA-34 | 69.1 | 69.2 | 69.0 | - | - |
| 2021--TIP | D2CNet | VGG-16 | 70.0 | 74.1 | 66.2 | 11.3/50.2/67.8/74.5/69.5/76.5 | 58.3 |
| 2020--arXiv | IIM | VGG-16 | 73.2 | 77.9 | 69.2 | 10.1/44.1/70.7/82.4/83.0/61.4 | 58.7 |
| 2020--arXiv | IIM | HRNet | 76.2 | 81.3 | 71.7 | 12.0/46.0/73.2/85.5/86.7/64.3 | 61.3 |