Crowd Counting with Density Adaption Networks
This work addresses the challenge of uneven crowd distribution for surveillance applications, representing an incremental improvement in domain-specific crowd counting.
The paper tackles the problem of accurate crowd counting in dynamic surveillance scenarios by proposing a lightweight deep learning framework that adapts to different crowd density levels, achieving promising improvements over state-of-the-art methods on datasets like UCF_CC_50 and ShanghaiTech.
Crowd counting is one of the core tasks in various surveillance applications. A practical system involves estimating accurate head counts in dynamic scenarios under different lightning, camera perspective and occlusion states. Previous approaches estimate head counts despite that they can vary dramatically in different density settings; the crowd is often unevenly distributed and the results are therefore unsatisfactory. In this paper, we propose a lightweight deep learning framework that can automatically estimate the crowd density level and adaptively choose between different counter networks that are explicitly trained for different density domains. Experiments on two recent crowd counting datasets, UCF_CC_50 and ShanghaiTech, show that the proposed mechanism achieves promising improvements over state-of-the-art methods. Moreover, runtime speed is 20 FPS on a single GPU.