Counting Crowds in Bad Weather
This addresses a domain-specific problem for computer vision applications in surveillance or public safety, but it is incremental as it builds on existing crowd counting methods by adapting them to weather variations.
The paper tackles the problem of crowd counting under adverse weather conditions like haze, rain, and snow, where existing methods fail due to appearance variations, and proposes a model that learns weather queries to handle these degradations, showing effectiveness on benchmark datasets.
Crowd counting has recently attracted significant attention in the field of computer vision due to its wide applications to image understanding. Numerous methods have been proposed and achieved state-of-the-art performance for real-world tasks. However, existing approaches do not perform well under adverse weather such as haze, rain, and snow since the visual appearances of crowds in such scenes are drastically different from those images in clear weather of typical datasets. In this paper, we propose a method for robust crowd counting in adverse weather scenarios. Instead of using a two-stage approach that involves image restoration and crowd counting modules, our model learns effective features and adaptive queries to account for large appearance variations. With these weather queries, the proposed model can learn the weather information according to the degradation of the input image and optimize with the crowd counting module simultaneously. Experimental results show that the proposed algorithm is effective in counting crowds under different weather types on benchmark datasets. The source code and trained models will be made available to the public.