Awesome-Crowd-Counting

Why we included this project

Anyone working on crowd counting, density estimation, or pedestrian analysis will find this a solid starting point. The author has grouped the literature by method family, so you can jump straight to the papers that match your constraints, whether that means lightweight models, transformer-based approaches, or semi-supervised and transfer-learning work. It also tracks the major public datasets such as NWPU-Crowd, JHU-CROWD++, and ShanghaiTech, with links, and keeps a performance leaderboard that separates published results from unreviewed arXiv preprints. The code section points to usable implementations like the C^3 PyTorch framework and labeling tools, and the related-tasks links lead toward crowd localization and video surveillance. It is a curated index rather than runnable software, but for scoping a project or writing a related-work section it saves days of searching.