Spatiotemporal Dilated Convolution with Uncertain Matching for Video-based Crowd Estimation

TL;DR

This paper introduces STDNet, a spatiotemporal dense network for video crowd estimation that combines dilated convolutions with attention mechanisms and a patch-wise loss to improve accuracy and efficiency.

Contribution

The paper presents a novel STDNet architecture with 3D dilated convolutions and a patch-wise regression loss, addressing model size and label noise issues in video crowd counting.

Findings

Outperforms state-of-the-art methods on UCSD, Mall, and WorldExpo'10 datasets.

Effectively handles label noise with the proposed patch-wise regression loss.

Reduces model size while maintaining high accuracy.

Abstract

In this paper, we propose a novel SpatioTemporal convolutional Dense Network (STDNet) to address the video-based crowd counting problem, which contains the decomposition of 3D convolution and the 3D spatiotemporal dilated dense convolution to alleviate the rapid growth of the model size caused by the Conv3D layer. Moreover, since the dilated convolution extracts the multiscale features, we combine the dilated convolution with the channel attention block to enhance the feature representations. Due to the error that occurs from the difficulty of labeling crowds, especially for videos, imprecise or standard-inconsistent labels may lead to poor convergence for the model. To address this issue, we further propose a new patch-wise regression loss (PRL) to improve the original pixel-wise loss. Experimental results on three video-based benchmarks, i.e., the UCSD, Mall and WorldExpo'10 datasets, show that STDNet outperforms both image- and video-based state-of-the-art methods. The source codes are released at \url{https://github.com/STDNet/STDNet}.