DENet: A Universal Network for Counting Crowd with Varying Densities and Scales

TL;DR

DENet is a hybrid network combining detection and density estimation to accurately count crowds with varying scales and densities, outperforming existing methods on multiple datasets.

Contribution

This paper introduces DENet, a novel hybrid network that integrates detection and density estimation for improved crowd counting accuracy.

Findings

Lower MAE on ShanghaiTech Part A, UCF, and WorldExpo'10 datasets.

Effective combination of detection and density estimation components.

Outperforms state-of-the-art methods in crowd counting accuracy.

Abstract

Counting people or objects with significantly varying scales and densities has attracted much interest from the research community and yet it remains an open problem. In this paper, we propose a simple but an efficient and effective network, named DENet, which is composed of two components, i.e., a detection network (DNet) and an encoder-decoder estimation network (ENet). We first run DNet on an input image to detect and count individuals who can be segmented clearly. Then, ENet is utilized to estimate the density maps of the remaining areas, where the numbers of individuals cannot be detected. We propose a modified Xception as an encoder for feature extraction and a combination of dilated convolution and transposed convolution as a decoder. In the ShanghaiTech Part A, UCF and WorldExpo'10 datasets, our DENet achieves lower Mean Absolute Error (MAE) than those of the state-of-the-art methods.