Composition Loss for Counting, Density Map Estimation and Localization in Dense Crowds

Abstract

With multiple crowd gatherings of millions of people every year in events ranging from pilgrimages to protests, concerts to marathons, and festivals to funerals; visual crowd analysis is emerging as a new frontier in computer vision. In particular, counting in highly dense crowds is a challenging problem with far-reaching applicability in crowd safety and management, as well as gauging political significance of protests and demonstrations. In this paper, we propose a novel approach that simultaneously solves the problems of counting, density map estimation and localization of people in a given dense crowd image. Our formulation is based on an important observation that the three problems are inherently related to each other making the loss function for optimizing a deep CNN decomposable. Since localization requires high-quality images and annotations, we introduce UCF-QNRF dataset that overcomes the shortcomings of previous datasets, and contains 1.25 million humans manually marked with dot annotations. Finally, we present evaluation measures and comparison with recent deep CNN networks, including those developed specifically for crowd counting. Our approach significantly outperforms state-of-the-art on the new dataset, which is the most challenging dataset with the largest number of crowd annotations in the most diverse set of scenes.

References in corpus (1)

• Densely Connected Convolutional Networks

Cited by in corpus (23)

• Crowd Counting and Density Estimation by Trellis Encoder-Decoder Network
• Dual Path Multi-Scale Fusion Networks with Attention for Crowd Counting
• Learning from Synthetic Data for Crowd Counting in the Wild
• Crowd Counting with Deep Structured Scale Integration Network
• C^3 Framework: An Open-source PyTorch Code for Crowd Counting
• Perspective-Guided Convolution Networks for Crowd Counting
• Ambient Sound Helps: Audiovisual Crowd Counting in Extreme Conditions
• Learn to Scale: Generating Multipolar Normalized Density Maps for Crowd Counting
• Feature-aware Adaptation and Density Alignment for Crowd Counting in Video Surveillance
• Multi-Resolution Fusion and Multi-scale Input Priors Based Crowd Counting
• Towards Using Count-level Weak Supervision for Crowd Counting
• LDC-Net: A Unified Framework for Localization, Detection and Counting in Dense Crowds
• Crowd Transformer Network
• NAS-Count: Counting-by-Density with Neural Architecture Search
• Attribute-aware Pedestrian Detection in a Crowd
• Improving Dense Crowd Counting Convolutional Neural Networks using Inverse k-Nearest Neighbor Maps and Multiscale Upsampling
• PCC Net: Perspective Crowd Counting via Spatial Convolutional Network
• Multi-Level Attentive Convoluntional Neural Network for Crowd Counting
• Efficient Crowd Counting via Structured Knowledge Transfer
• PANet: Perspective-Aware Network with Dynamic Receptive Fields and Self-Distilling Supervision for Crowd Counting
• Wisdom of (Binned) Crowds: A Bayesian Stratification Paradigm for Crowd Counting
• Relevant Region Prediction for Crowd Counting
• Recurrent Distillation based Crowd Counting