Recent Deep Learning in Crowd Behaviour Analysis: A Brief Review

2025-05-23

AI Summary

This survey addresses the lack of cross-paradigm comparison and integration of physical priors in deep learning–based crowd behavior analysis. We propose a unified analytical framework that systematically integrates purely data-driven models (CNNs, RNNs, GNNs, Transformers) with physics-guided approaches (Physics-Informed Neural Networks, multi-agent simulations, spatiotemporal graph learning), explicitly clarifying technical evolution and identifying critical research gaps. Through rigorous comparative evaluation across behavior recognition and prediction tasks, we characterize performance boundaries and domain-specific applicability of each paradigm. Our analysis yields three concrete future directions: enhancing model interpretability, improving few-shot generalization, and achieving robust long-horizon forecasting. The work serves as a structured entry point for researchers and offers a forward-looking perspective grounded in both empirical assessment and theoretical synthesis.

Technology Category

• Computer Vision: Low Level & Physics-based Vision
• Humans and AI: Human-Aware Planning and Behavior Prediction
• Game Theory and Economic Paradigms: Behavioral Game Theory

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

Crowd behaviour analysis is essential to numerous real-world applications, such as public safety and urban planning, and therefore has been studied for decades. In the last decade or so, the development of deep learning has significantly propelled the research on crowd behaviours. This chapter reviews recent advances in crowd behaviour analysis using deep learning. We mainly review the research in two core tasks in this field, crowd behaviour prediction and recognition. We broadly cover how different deep neural networks, after first being proposed in machine learning, are applied to analysing crowd behaviours. This includes pure deep neural network models as well as recent development of methodologies combining physics with deep learning. In addition, representative studies are discussed and compared in detail. Finally, we discuss the effectiveness of existing methods and future research directions in this rapidly evolving field. This chapter aims to provide a high-level summary of the ongoing deep learning research in crowd behaviour analysis. It intends to help new researchers who just entered this field to obtain an overall understanding of the ongoing research, as well as to provide a retrospective analysis for existing researchers to identify possible future directions

Problem

• Review deep learning advances in crowd behavior analysis
• Compare methods for crowd behavior prediction and recognition
• Discuss effectiveness and future research directions

Innovation

• Deep neural networks for crowd analysis
• Physics combined with deep learning
• Behavior prediction and recognition models