A Deep Learning-Based Crowd Counting Method and System Implementation on Neural Processing Unit Platform
Source abstract
Abstract
In this paper, a deep learning-based method is proposed for crowd-counting problems. Specifically, by utilizing the convolution kernel density map, the ground truth is generated dynamically to enhance the feature-extracting ability of the generator model. Meanwhile, the “cross stage partial” module is integrated into congested scene recognition network (CSRNet) to obtain a lightweight network model. In addition, to compensate for the accuracy drop owing to the lightweight model, we take advantage of “structured knowledge transfer” to train the model in an end-to-end manner. It aims to accelerate the fitting speed and enhance the learning ability of the student model. The crowd-counting system solution for edge computing is also proposed and implemented on an embedded device equipped with a neural processing unit. Simulations demonstrate the performance improvement of the proposed solution in terms of model size, processing speed and accuracy. The performance on the Venice dataset shows that the mean absolute error (MAE) and the root mean squared error (RMSE) of our model drop by 32.63% and 39.18% compared with CSRNet. Meanwhile, the performance on the ShanghaiTech PartB dataset reveals that the MAE and the RMSE of our model are close to those of CSRNet. Therefore, we provide a novel embedded platform system scheme for public safety pre-warning applications.
Source: openalex_inverted_index · Quality: baseline
Research context
OpenAlex topics: Video Surveillance and Tracking Methods · Anomaly Detection Techniques and Applications · Fire Detection and Safety Systems
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
References (48)
Show source-provided references
1. A. Bochkovskiy, C. Wang and H. M. Liao, “Yolov4: Optimal speed and accuracy of object detection,” arXiv preprint arXiv:2004.10934, 2020.
2. SSD: Single shot multibox detector
3. Single-image crowd counting via multi-column convolutional neural network
4. CSRNet: Dilated convolutional neural networks for understanding the highly congested scenes
5. Perspective-guided convolution networks for crowd counting
6. Encoder decoder based convolutional neural networks with multi-scale-aware modules for crowd counting
7. L. Zhu, Z. Zhao, C. Lu, Y. Lin, Y. Peng et al., “Dual path multi-scale fusion networks with attention for crowd counting,” arXiv preprint, arXiv:1902.01115, 2019.
8. Deeplab: Semantic image segmentation with deep convolutional nets, atrous convolution, and fully connected CRFs10.1109/TPAMI.2017.2699184 ↗
9. Context-aware crowd counting
10. Adaptive dilated network with self-correction supervision for counting
11. To choose or to fuse? Scale selection for crowd counting
12. Generating high-quality crowd density maps using contextual pyramid CNNs
13. M. M. Oghaz, A. R. Khadka, V. Argyriou and P. Remagnino, “Content aware density map for crowd counting and density estimation,” arXiv preprint arXiv:1906.07258, 2019.
14. Adaptive density map generation for crowd counting
15. Kernel-based density map generation for dense object counting10.1109/TPAMI.2020.3022878 ↗
16. G. Hinton, O. Vinyals and J. Dean, “Distilling the knowledge in a neural network,” arXiv preprint, arXiv:1503.02531, 2015.
17. Fast human pose estimation
18. Mobilecount: An efficient encoderdecoder framework for real-time crowd counting10.1016/j.neucom.2020.05.056 ↗
19. Efficient crowd counting via structured knowledge transfer
20. Shufflecount: Task-specific knowledge distillation for crowd counting
21. CSPNet: A new backbone that can enhance learning capability of CNN
22. ECA-net: Efficient channel attention for deep convolutional neural networks
23. K. Simonyan and A. Zisserman, “Very deep convolutional networks for large-scale image recognition,” arXiv preprint, arXiv:1409.1556, 2014.
24. A deeply-recursive convolutional network for crowd counting
25. Switching convolutional neural network for crowd counting
26. DFE-crowd: Dense feature extraction for single image crowd counting
27. AMS-CNN: Attentive multi-stream CNN for video based crowd counting10.1007/s13735-021-00220-7 ↗
28. Decidenet: Counting varying density crowds through attention guided detection and density estimation
29. Geometric and physical constraints for drone-based head plane crowd density estimation
30. PhNet: Parasite-host network for video crowd counting
31. An effective modular approach for crowd counting in an image using convolutional neural networks
32. Hadf-crowd: A hierarchical attention-based dense feature extraction network for single-image crowd counting10.3390/s21103483 ↗
33. M. Marsden, K. McGuinness, S. Little and N. E. O’Connor, “Fully convolutional crowd counting on highly congested scenes,” arXiv preprint arXiv:1612.00220, 2016.10.5220/0006097300270033 ↗
34. Deep residual learning for image recognition
35. A crowd counting framework combining with crowd location10.1155/2021/8793101 ↗
36. Crowd density estimation using fusion of multi-layer features10.1109/TITS.2020.2983475 ↗
37. Learning a deep network with cross-hierarchy aggregation for crowd counting10.1016/j.knosys.2020.106691 ↗
38. Crowd counting via scale-adaptive convolutional neural network
39. Leveraging unlabeled data for crowd counting by learning to rank
40. R. Bahmanyar, E. Vig and P. Reinartz, “MRCNet: Crowd counting and density map estimation in aerial and ground imagery,” arXiv preprint arXiv:1909.12743, 2019.
41. Bayesian loss for crowd count estimation with point supervision
42. Distribution matching for crowd counting
43. Crowd counting with deep structured scale integration network
44. Scale aggregation network for accurate and efficient crowd counting
45. CNN-based cascaded multi-task learning of high-level prior and density estimation for crowd counting
46. A real-time UAV crowd counting system based on edge computing
47. People counting on low cost embedded hardware during the sars-cov-2 pandemic
48. Searching for mobilenetv3