Crowd Counting Using Scale-Aware Attention Networks

Abstract

In this paper, we consider the problem of crowd counting in images. Given an image of a crowded scene, our goal is to estimate the density map of this image, where each pixel value in the density map corresponds to the crowd density at the corresponding location in the image. Given the estimated density map, the final crowd count can be obtained by summing over all values in the density map. One challenge of crowd counting is the scale variation in images. In this work, we propose a novel scale-aware attention network to address this challenge. Using the attention mechanism popular in recent deep learning architectures, our model can automatically focus on certain global and local scales appropriate for the image. By combining these global and local scale attention, our model outperforms other state-of-the-art methods for crowd counting on several benchmark datasets.

References in corpus (6)

• Show, Attend and Tell: Neural Image Caption Generation with Visual Attention
• R-FCN: Object Detection via Region-based Fully Convolutional Networks
• CrowdNet: A Deep Convolutional Network for Dense Crowd Counting
• Ten Years of Pedestrian Detection, What Have We Learned?
• Mixture of Counting CNNs: Adaptive Integration of CNNs Specialized to Specific Appearance for Crowd Counting
• Visual Translation Embedding Network for Visual Relation Detection

Cited by in corpus (3)

• Crowd counting with crowd attention convolutional neural network
• Dual Path Multi-Scale Fusion Networks with Attention for Crowd Counting
• Multi-Level Attentive Convoluntional Neural Network for Crowd Counting