Count2Density: Crowd density estimation without location-level annotations

Mattia Litrico, Feng Chen, Michael Pound, Sotirios A. Tsaftaris, Sebastiano Battiato, Mario Valerio Giuffrida

Pattern Recognition (2026)

Litrico, M., Chen, F., Pound, M., Tsaftaris, S.A., Battiato, S., Giuffrida, M. V. (2026). Count2Density: Crowd density estimation without location-level annotations. In Pattern Recognition (113755).

wp-content/uploads/2020/10/tex.png
Get Paper
@article{LITRICO2026113755,
title = {Count2Density: Crowd density estimation without location-level annotations},
journal = {Pattern Recognition},
pages = {113755},
year = {2026},
issn = {0031-3203},
doi = {https://doi.org/10.1016/j.patcog.2026.113755},
url = {https://www.sciencedirect.com/science/article/pii/S003132032600720X},
author = {Mattia Litrico and Feng Chen and Michael Pound and Sotirios A. Tsaftaris and Sebastiano Battiato and Mario Valerio Giuffrida},
keywords = {Density estimation, Count-level annotations, Weak-supervision},
abstract = {Crowd density estimation is a well-known computer vision task aimed at estimating the density distribution of people in an image. The primary challenge in this domain is the reliance on fine-grained location-level annotations – i.e., points placed on top of each individual—to train deep networks. Collecting such detailed annotations is both tedious, time-consuming, and poses a significant barrier to scalability for real-world applications. To alleviate this burden, we present Count2Density: a novel pipeline designed to predict meaningful density maps containing quantitative spatial information using only count-level annotations (i.e., the total number of people) during training. To achieve this, Count2Density generates pseudo-density maps leveraging past predictions stored in a Historical Map Bank, thereby reducing confirmation bias. This bank is initialised using an unsupervised saliency estimator to provide an initial spatial prior and is iteratively updated with an Exponential Moving Average of predicted density maps. These pseudo-density maps are obtained by sampling locations from estimated crowd areas using a hypergeometric distribution, with the number of samplings determined by the count-level annotations. To further enhance the spatial awareness of the model and promote robust feature learning, we add a self-supervised contrastive spatial regulariser to encourage similar feature representations within crowded regions while maximising dissimilarity with background regions. Experimental results demonstrate that our approach significantly outperforms cross-domain adaptation methods and achieves better results than recent state-of-the-art approaches in semi-supervised settings across several datasets. Additional analyses validate the effectiveness of each individual component of our pipeline, including the self-supervised contrastive regulariser, confirming the ability of Count2Density to effectively retrieve spatial information from count-level annotations and enabling accurate subregion counting.}
}

Abstract

Crowd density estimation is a well-known computer vision task aimed at estimating the density distribution of people in an image. The primary challenge in this domain is the reliance on fine-grained location-level annotations – i.e., points placed on top of each individual—to train deep networks. Collecting such detailed annotations is both tedious, time-consuming, and poses a significant barrier to scalability for real-world applications. To alleviate this burden, we present Count2Density: a novel pipeline designed to predict meaningful density maps containing quantitative spatial information using only count-level annotations (i.e., the total number of people) during training. To achieve this, Count2Density generates pseudo-density maps leveraging past predictions stored in a Historical Map Bank, thereby reducing confirmation bias. This bank is initialised using an unsupervised saliency estimator to provide an initial spatial prior and is iteratively updated with an Exponential Moving Average of predicted density maps. These pseudo-density maps are obtained by sampling locations from estimated crowd areas using a hypergeometric distribution, with the number of samplings determined by the count-level annotations. To further enhance the spatial awareness of the model and promote robust feature learning, we add a self-supervised contrastive spatial regulariser to encourage similar feature representations within crowded regions while maximising dissimilarity with background regions. Experimental results demonstrate that our approach significantly outperforms cross-domain adaptation methods and achieves better results than recent state-of-the-art approaches in semi-supervised settings across several datasets. Additional analyses validate the effectiveness of each individual component of our pipeline, including the self-supervised contrastive regulariser, confirming the ability of Count2Density to effectively retrieve spatial information from count-level annotations and enabling accurate subregion counting.