Li Pengfei, Zhang Min, Wan Jian, Jiang Ming
Computer & Software School, Hangzhou Dianzi University, Hangzhou, Zhejiang, China.
PeerJ Comput Sci. 2022 Mar 18;8:e902. doi: 10.7717/peerj-cs.902. eCollection 2022.
Crowd counting has been widely studied by deep learning in recent years. However, due to scale variation caused by perspective distortion, crowd counting is still a challenging task. In this paper, we propose a Densely Connected Multi-scale Pyramid Network (DMPNet) for count estimation and the generation of high-quality density maps. The key component of our network is the Multi-scale Pyramid Network (MPN), which can extract multi-scale features of the crowd effectively while keeping the resolution of the input feature map and the number of channels unchanged. To increase the information transfer between the network layer, we used dense connections to connect multiple MPNs. In addition, we also designed a novel loss function, which can help our model achieve better convergence. To evaluate our method, we conducted extensive experiments on three challenging benchmark crowd counting datasets. Experimental results show that compared with the state-of-the-art algorithms, DMPNet performs well in both parameters and results. The code is available at: https://github.com/lpfworld/DMPNet.
近年来，深度学习对人群计数进行了广泛研究。然而，由于视角扭曲导致的尺度变化，人群计数仍然是一项具有挑战性的任务。在本文中，我们提出了一种用于计数估计和生成高质量密度图的密集连接多尺度金字塔网络（DMPNet）。我们网络的关键组件是多尺度金字塔网络（MPN），它可以有效地提取人群的多尺度特征，同时保持输入特征图的分辨率和通道数不变。为了增加网络层之间的信息传递，我们使用密集连接来连接多个MPN。此外，我们还设计了一种新颖的损失函数，它可以帮助我们的模型实现更好的收敛。为了评估我们的方法，我们在三个具有挑战性的基准人群计数数据集上进行了广泛的实验。实验结果表明，与现有最先进算法相比，DMPNet在参数和结果方面均表现出色。代码可在以下网址获取：https://github.com/lpfworld/DMPNet。
