ZIP: Scalable Crowd Counting via Zero-Inflated Poisson Modeling
This work improves crowd counting accuracy for applications like surveillance and event management, representing a strong specific gain rather than a broad paradigm shift.
The paper tackled the problem of crowd counting by addressing limitations of existing methods that use MSE losses, which dilute supervision due to sparse annotations and mismatch discrete count data, and introduced ZIP, a framework using Zero-Inflated Poisson modeling that consistently outperformed state-of-the-art methods across multiple benchmarks.
Most crowd counting methods directly regress blockwise density maps using Mean Squared Error (MSE) losses. This practice has two key limitations: (1) it fails to account for the extreme spatial sparsity of annotations - over 95% of 8x8 blocks are empty across standard benchmarks, so supervision signals in informative regions are diluted by the predominant zeros; (2) MSE corresponds to a Gaussian error model that poorly matches discrete, non-negative count data. To address these issues, we introduce ZIP, a scalable crowd counting framework that models blockwise counts with a Zero-Inflated Poisson likelihood: a zero-inflation term learns the probability a block is structurally empty (handling excess zeros), while the Poisson component captures expected counts when people are present (respecting discreteness). We provide a generalization analysis showing a tighter risk bound for ZIP than MSE-based losses and DMCount provided that the training resolution is moderately large. To assess the scalability of ZIP, we instantiate it on backbones spanning over 100x in parameters/compute. Experiments on ShanghaiTech A & B, UCF-QNRF, and NWPU-Crowd demonstrate that ZIP consistently surpasses state-of-the-art methods across all model scales.