RepSFNet : A Single Fusion Network with Structural Reparameterization for Crowd Counting

TL;DR

RepSFNet is a lightweight, real-time crowd counting network that uses reparameterized kernels and feature fusion modules to improve accuracy and efficiency in variable-density scenes.

Contribution

The paper introduces RepSFNet, a novel crowd counting architecture with reparameterized kernels and fusion modules, reducing complexity while maintaining high accuracy.

Findings

Achieves competitive accuracy on multiple datasets.

Reduces inference latency by up to 34%.

Operates effectively in real-time and low-power environments.

Abstract

Crowd counting remains challenging in variable-density scenes due to scale variations, occlusions, and the high computational cost of existing models. To address these issues, we propose RepSFNet (Reparameterized Single Fusion Network), a lightweight architecture designed for accurate and real-time crowd estimation. RepSFNet leverages a RepLK-ViT backbone with large reparameterized kernels for efficient multi-scale feature extraction. It further integrates a Feature Fusion module combining Atrous Spatial Pyramid Pooling (ASPP) and Context-Aware Network (CAN) to achieve robust, density-adaptive context modeling. A Concatenate Fusion module is employed to preserve spatial resolution and generate high-quality density maps. By avoiding attention mechanisms and multi-branch designs, RepSFNet significantly reduces parameters and computational complexity. The training objective combines Mean Squared Error and Optimal Transport loss to improve both count accuracy and spatial distribution alignment. Experiments conducted on ShanghaiTech, NWPU, and UCF-QNRF datasets demonstrate that RepSFNet achieves competitive accuracy while reducing inference latency by up to 34 percent compared to recent state-of-the-art methods, making it suitable for real-time and low-power edge computing applications.