Problem
Research questions and friction points this paper is trying to address.
image denoising
lookup table (LUT)
memory efficiency
color distortion
edge devices
Hybrid-LUT: Channel-Aware Hybrid Lookup Table and Filtering for Efficient Image Denoising
🤖 AI Summary
This work addresses the high SRAM overhead of existing RGB lookup table (LUT)-based image denoising methods, which process all three color channels in parallel, and the color distortion issues arising from denoising only the Y channel in YUV space. To overcome these limitations, the authors propose an asymmetric channel processing framework in the YUV domain: the Y channel is restored using a multi-band LUT combined with pixel-wise weight fusion to preserve texture details, while lightweight filtering is applied to the UV channels to maintain color fidelity. This approach uniquely integrates LUTs and filtering within an asymmetric YUV optimization strategy, reducing LUT memory usage by two-thirds to just 421 KB without compromising real-time throughput. Evaluated on real-world images, the method achieves state-of-the-art performance, improving CPSNR by at least 0.63 dB over current LUT-based techniques.
📝 Abstract
Lookup table (LUT)-based image denoising methods have attracted increasing attention due to their high efficiency and hardware-friendly properties. However, existing RGB-LUT approaches require three identical LUTs to process RGB channels in parallel, resulting in large on-chip SRAM consumption. A simple alternative is to apply LUT processing only to the luminance (Y) channel in the YUV color space to reduce memory usage. However, this naive strategy leads to degraded restoration quality, since ignoring the chrominance (UV) channels introduces color distortion and residual artifacts. In this work, we propose Hybrid-LUT, a YUV-based asymmetric channel-processing framework that combines LUT and filtering in a unified design. Specifically, a multi-band LUT branch with pixel-level weight fusion is applied to the Y channel to recover fine textures, while lightweight filtering is used for the UV channels to maintain color consistency. This design reduces LUT storage by two-thirds compared with RGB-LUT methods while maintaining the same runtime throughput. Extensive experiments show that Hybrid-LUT achieves state-of-the-art (SOTA) performance across multiple benchmarks with only 421 KB of storage. In particular, our method surpasses existing LUT-based denoising approaches by at least 0.63 dB CPSNR on real-world datasets, demonstrating its effectiveness for image denoising on resource-constrained edge devices. The project is available at https://github.com/Ai-ZL/Hybrid-LUT .
Innovation
Methods, ideas, or system contributions that make the work stand out.
Hybrid-LUT
channel-aware processing
lookup table (LUT)
image denoising
edge computing
🔎 Similar Papers
No similar papers found.
