Multi-Modal Hyper-Graph Fusion for Low-Light Crowd Counting

2026-06-16

AI Summary

This work addresses the significant performance degradation of crowd counting under low-light conditions, where RGB-only representations often fail. To overcome this limitation, the authors propose LCNet, the first approach to incorporate depth and Canny edge maps as geometric and structural priors. LCNet employs a multimodal hypergraph fusion module to model high-order complementary relationships among RGB, depth, and edge modalities, combined with a deformable rectangular sparse attention mechanism for efficient dense prediction. The study contributes the first real-world low-light crowd dataset, LC-Crowd, along with two synthetic benchmarks. Extensive experiments demonstrate that LCNet substantially outperforms existing state-of-the-art methods across all three datasets, confirming its robustness and accuracy under both extreme and non-uniform low-light conditions.

Abstract

Crowd counting is a fundamental task in computer vision. However, crowd counting in low-light environments remains largely underexplored, despite its practical importance in the real world. Existing methods mainly focus on well-lit scenes or rely on single-modality Red-Green-Blue (RGB) representations, which often become unreliable under extreme darkness and complex non-uniform illumination. To handle this problem, we construct three new low-light crowd counting benchmarks, which consist of two synthetic datasets, SHA\_Dark and SHB\_Dark, and a real-world benchmark LC-Crowd (Low-light Crowd Dataset). Inspired by Retinex-based physical modeling, we introduce depth and Canny edge cues as complementary geometric and structural priors to enhance the intrinsic reflectance representation under low-light conditions. We propose a Multi-Modal Hyper-Graph Fusion module, which formulates RGB appearance, depth geometry, and edge structure cues as nodes in a unified hyper-graph and explicitly captures their high-order complementary relationships via dynamic hyperedge construction and message passing. Furthermore, to adaptively allocate computation in dense prediction, we propose a Deformable Rectangular Sparse Attention (DRSA) module, which concentrates computation on informative regions through anchor-aware estimation and adaptive rectangular window modeling. Based on these designs, we develop a unified Low-Light Counting Network (LCNet) for robust low-light crowd counting. Extensive experiments on three benchmarks demonstrate that the proposed method achieves the best overall performance against existing state-of-the-art (SOTA) methods. The code is in the supplementary material. The datasets will be made public upon acceptance.

Problem

• low-light crowd counting
• multi-modal fusion
• RGB-depth-edge
• non-uniform illumination
• crowd density estimation

Innovation

• Multi-Modal Hyper-Graph Fusion
• Low-Light Crowd Counting
• Deformable Rectangular Sparse Attention
• Retinex-based Modeling
• Depth and Edge Priors