WREN network enhances low-light images using Retinex theory and U-Nets

Researchers have introduced WREN, a novel neural network designed for low-light image enhancement. WREN utilizes a Retinex theory-based approach, decomposing images into reflectance and illumination maps using a double U-Net-like structure. A subsequent network, incorporating a Transformer block, specifically enhances the illumination map. The model is trained end-to-end with a scale-invariant loss function to ensure robustness across various dynamic range scenes and lighting conditions, achieving state-of-the-art performance on multiple datasets. AI

IMPACT Introduces a novel architecture for low-light image enhancement, potentially improving performance in computer vision tasks under challenging lighting conditions.

RANK_REASON This is a research paper detailing a new neural network architecture for image enhancement. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.CV →

• arXiv
• Retinex theory
• Transformer++
• U-Net
• WREN

• paper
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →