HA-CCN: Hierarchical Attention-based Crowd Counting Network

Abstract

Single image-based crowd counting has recently witnessed increased focus, but many leading methods are far from optimal, especially in highly congested scenes. In this paper, we present Hierarchical Attention-based Crowd Counting Network (HA-CCN) that employs attention mechanisms at various levels to selectively enhance the features of the network. The proposed method, which is based on the VGG16 network, consists of a spatial attention module (SAM) and a set of global attention modules (GAM). SAM enhances low-level features in the network by infusing spatial segmentation information, whereas the GAM focuses on enhancing channel-wise information in the higher level layers. The proposed method is a single-step training framework, simple to implement and achieves state-of-the-art results on different datasets. Furthermore, we extend the proposed counting network by introducing a novel set-up to adapt the network to different scenes and datasets via weak supervision using image-level labels. This new set up reduces the burden of acquiring labour intensive point-wise annotations for new datasets while improving the cross-dataset performance.

Accepted for publication at IEEE Transactions on Image Processing (TIP) 2019

References in corpus (4)

• A Survey of Recent Advances in CNN-based Single Image Crowd Counting and Density Estimation
• Crowded Scene Analysis: A Survey
• Few-Example Object Detection with Model Communication
• Learning Short-Cut Connections for Object Counting

Cited by in corpus (13)

• A General Survey on Attention Mechanisms in Deep Learning
• A Self-Training Approach for Point-Supervised Object Detection and Counting in Crowds
• Neuron Linear Transformation: Modeling the Domain Shift for Crowd Counting
• Encoder-Decoder Based Convolutional Neural Networks with Multi-Scale-Aware Modules for Crowd Counting
• Spatiotemporal Dilated Convolution with Uncertain Matching for Video-based Crowd Estimation
• Crowd Counting via Perspective-Guided Fractional-Dilation Convolution
• Dilated-Scale-Aware Attention ConvNet For Multi-Class Object Counting
• TreeFormer: a Semi-Supervised Transformer-based Framework for Tree Counting from a Single High Resolution Image
• Video Crowd Localization with Multi-focus Gaussian Neighborhood Attention and a Large-Scale Benchmark
• Forget Less, Count Better: A Domain-Incremental Self-Distillation Learning Benchmark for Lifelong Crowd Counting
• Learning Discriminative Features for Crowd Counting
• BBA-net: A bi-branch attention network for crowd counting
• An Improved Normed-Deformable Convolution for Crowd Counting