Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

BiEvLight: Bi-level Learning of Task-Aware Event Refinement for Low-Light Image Enhancement

Zishu Yao, Xiang-Xiang Su, Shengning Zhou, Guang-Yong Chen, Guodong Fan, Xing Chen

TL;DR

This work proposes BiEvLight, a hierarchical and task-aware framework that collaboratively optimizes enhancement and denoising by exploiting their intrinsic interdependence and exploits the strong gradient correlation between images and events to build a gradient-guided event denoising prior that alleviates insufficient denoising in heavily noisy regions.

Abstract

Event cameras, with their high dynamic range, show great promise for Low-light Image Enhancement (LLIE). Existing works primarily focus on designing effective modal fusion strategies. However, a key challenge is the dual degradation from intrinsic background activity (BA) noise in events and low signal-to-noise ratio (SNR) in images, which causes severe noise coupling during modal fusion, creating a critical performance bottleneck. We therefore posit that precise event denoising is the prerequisite to unlocking the full potential of event-based fusion. To this end, we propose BiEvLight, a hierarchical and task-aware framework that collaboratively optimizes enhancement and denoising by exploiting their intrinsic interdependence. Specifically, BiEvLight exploits the strong gradient correlation between images and events to build a gradient-guided event denoising prior that alleviates insufficient denoising in heavily noisy regions. Moreover, instead of treating event denoising as a static pre-processing stage-which inevitably incurs a trade-off between over- and under-denoising and cannot adapt to the requirements of a specific enhancement objective-we recast it as a bilevel optimization problem constrained by the enhancement task. Through cross-task interaction, the upper-level denoising problem learns event representations tailored to the lower-level enhancement objective, thereby substantially improving overall enhancement quality. Extensive experiments on the Real-world noise Dataset SDE demonstrate that our method significantly outperforms state-of-the-art (SOTA) approaches, with average improvements of 1.30dB in PSNR, 2.03dB in PSNR* and 0.047 in SSIM, respectively. The code will be publicly available at https://github.com/iijjlk/BiEvlight.

BiEvLight: Bi-level Learning of Task-Aware Event Refinement for Low-Light Image Enhancement

TL;DR

This work proposes BiEvLight, a hierarchical and task-aware framework that collaboratively optimizes enhancement and denoising by exploiting their intrinsic interdependence and exploits the strong gradient correlation between images and events to build a gradient-guided event denoising prior that alleviates insufficient denoising in heavily noisy regions.

Abstract

Event cameras, with their high dynamic range, show great promise for Low-light Image Enhancement (LLIE). Existing works primarily focus on designing effective modal fusion strategies. However, a key challenge is the dual degradation from intrinsic background activity (BA) noise in events and low signal-to-noise ratio (SNR) in images, which causes severe noise coupling during modal fusion, creating a critical performance bottleneck. We therefore posit that precise event denoising is the prerequisite to unlocking the full potential of event-based fusion. To this end, we propose BiEvLight, a hierarchical and task-aware framework that collaboratively optimizes enhancement and denoising by exploiting their intrinsic interdependence. Specifically, BiEvLight exploits the strong gradient correlation between images and events to build a gradient-guided event denoising prior that alleviates insufficient denoising in heavily noisy regions. Moreover, instead of treating event denoising as a static pre-processing stage-which inevitably incurs a trade-off between over- and under-denoising and cannot adapt to the requirements of a specific enhancement objective-we recast it as a bilevel optimization problem constrained by the enhancement task. Through cross-task interaction, the upper-level denoising problem learns event representations tailored to the lower-level enhancement objective, thereby substantially improving overall enhancement quality. Extensive experiments on the Real-world noise Dataset SDE demonstrate that our method significantly outperforms state-of-the-art (SOTA) approaches, with average improvements of 1.30dB in PSNR, 2.03dB in PSNR* and 0.047 in SSIM, respectively. The code will be publicly available at https://github.com/iijjlk/BiEvlight.
Paper Structure (24 sections, 24 equations, 9 figures, 6 tables, 1 algorithm)

This paper contains 24 sections, 24 equations, 9 figures, 6 tables, 1 algorithm.

Table of Contents

  1. Introduction
  2. Related Works
  3. Low-light Image Enhancement
  4. Event denoising
  5. Method
  6. Problem Formulation
  7. Mechanism of Imaging via BA Noise
  8. Low-Light Imaging with Retinex
  9. Bridging Events and Image
  10. Overall Framework
  11. BiEvLight
  12. Spatially-adaptive Gradient-guided Denoising
  13. Bi-level Learning of Task-Aware
  14. Optimization Algorithm
  15. Loss Function
  16. ...and 9 more sections

Figures (9)

  • Figure 1: (a) Two modalities exhibiting different types of degradation in low-light environments. (b) The proposed strategy, we aim to utilize cross-modal interaction to guide denoising to achieve the event representation adapted to lower-level enhancement tasks. (c) Visual comparison with the current most representative methods, as well as visualization of events after denoising. (d) Quantitative and visual comparison with current mainstream methods, the proposed method achieves optimal results on mainstream datasets.
  • Figure 2: (a) Problems and challenges of existing methods, (b) The proposed BiEvLight framework.
  • Figure 3: Qualitative results on SDE-in dataset.
  • Figure 4: Qualitative results on SDE-out dataset.
  • Figure 5: Visualization of Ablation Results for Event Denoising
  • ...and 4 more figures