Bidirectional Image-Event Guided Low-Light Image Enhancement

2025-06-06

AI Summary

Under extreme low-light conditions, frame-based cameras suffer from insufficient dynamic range—causing detail loss and motion blur—while existing event-guided methods neglect global low-frequency noise induced by dynamic illumination and local structural discontinuities arising from event sparsity. To address these challenges, this paper proposes a dual-modal collaborative enhancement framework. Methodologically, it introduces the first bidirectional cross-attention fusion (BCAF) mechanism and an event feature enhancement (EFE) module incorporating high-pass frequency filtering, jointly preserving frame-based texture fidelity while improving structural continuity and illumination robustness of event data. Evaluated on our newly established high-quality RELIE dataset, the proposed method achieves a 0.96 dB PSNR gain and a 0.03 reduction in LPIPS, significantly outperforming state-of-the-art approaches.

Technology Category

• Computer Vision: Multi-modal Vision
• Intelligent Robots: Multimodal Perception & Sensor Fusion
• Machine Learning: Multimodal Learning

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Responsible Web: Data and user privacy-enhancing technologies for the Web

Abstract

Under extreme low-light conditions, traditional frame-based cameras, due to their limited dynamic range and temporal resolution, face detail loss and motion blur in captured images. To overcome this bottleneck, researchers have introduced event cameras and proposed event-guided low-light image enhancement algorithms. However, these methods neglect the influence of global low-frequency noise caused by dynamic lighting conditions and local structural discontinuities in sparse event data. To address these issues, we propose an innovative Bidirectional guided Low-light Image Enhancement framework (BiLIE). Specifically, to mitigate the significant low-frequency noise introduced by global illumination step changes, we introduce the frequency high-pass filtering-based Event Feature Enhancement (EFE) module at the event representation level to suppress the interference of low-frequency information, and preserve and highlight the high-frequency edges.Furthermore, we design a Bidirectional Cross Attention Fusion (BCAF) mechanism to acquire high-frequency structures and edges while suppressing structural discontinuities and local noise introduced by sparse event guidance, thereby generating smoother fused representations.Additionally, considering the poor visual quality and color bias in existing datasets, we provide a new dataset (RELIE), with high-quality ground truth through a reliable enhancement scheme. Extensive experimental results demonstrate that our proposed BiLIE outperforms state-of-the-art methods by 0.96dB in PSNR and 0.03 in LPIPS.

Problem

• Mitigates global low-frequency noise in low-light images
• Addresses local structural discontinuities in sparse event data
• Improves visual quality and color bias in datasets

Innovation

• Frequency high-pass filtering for noise suppression
• Bidirectional Cross Attention Fusion for smoother edges
• New dataset RELIE for reliable enhancement