Learning Discriminative Features for Crowd Counting

TL;DR

This paper introduces a discriminative feature learning framework for crowd counting that enhances localization and foreground-background discrimination using masked feature prediction and contrastive learning modules, applicable to dense scene analysis.

Contribution

It proposes a novel plug-and-play framework with masked feature prediction and contrastive learning modules to improve crowd counting accuracy in congested scenes.

Findings

Improved crowd counting accuracy in highly congested scenes.

Enhanced localization ability through masked feature prediction.

Better discrimination between foreground and background objects.

Abstract

Crowd counting models in highly congested areas confront two main challenges: weak localization ability and difficulty in differentiating between foreground and background, leading to inaccurate estimations. The reason is that objects in highly congested areas are normally small and high level features extracted by convolutional neural networks are less discriminative to represent small objects. To address these problems, we propose a learning discriminative features framework for crowd counting, which is composed of a masked feature prediction module (MPM) and a supervised pixel-level contrastive learning module (CLM). The MPM randomly masks feature vectors in the feature map and then reconstructs them, allowing the model to learn about what is present in the masked regions and improving the model's ability to localize objects in high density regions. The CLM pulls targets close to each other and pushes them far away from background in the feature space, enabling the model to discriminate foreground objects from background. Additionally, the proposed modules can be beneficial in various computer vision tasks, such as crowd counting and object detection, where dense scenes or cluttered environments pose challenges to accurate localization. The proposed two modules are plug-and-play, incorporating the proposed modules into existing models can potentially boost their performance in these scenarios.