Interlayer and Intralayer Scale Aggregation for Scale-invariant Crowd Counting

TL;DR

This paper introduces a single-column network for crowd counting that effectively captures multi-scale features and maintains scale-invariance, outperforming multi-column methods in accuracy and transferability.

Contribution

The paper proposes a novel single-column network with interlayer and intralayer scale-invariant features, improving over multi-column approaches in crowd counting tasks.

Findings

Outperforms state-of-the-art methods in accuracy

Demonstrates strong transferability across datasets

Achieves scale-invariance in crowd counting

Abstract

Crowd counting is an important vision task, which faces challenges on continuous scale variation within a given scene and huge density shift both within and across images. These challenges are typically addressed using multi-column structures in existing methods. However, such an approach does not provide consistent improvement and transferability due to limited ability in capturing multi-scale features, sensitiveness to large density shift, and difficulty in training multi-branch models. To overcome these limitations, a Single-column Scale-invariant Network (ScSiNet) is presented in this paper, which extracts sophisticated scale-invariant features via the combination of interlayer multi-scale integration and a novel intralayer scale-invariant transformation (SiT). Furthermore, in order to enlarge the diversity of densities, a randomly integrated loss is presented for training our single-branch method. Extensive experiments on public datasets demonstrate that the proposed method consistently outperforms state-of-the-art approaches in counting accuracy and achieves remarkable transferability and scale-invariant property.