WaFIR improves image repair by routing parts by frequency
WaFIR: Wavelet-Fourier Mixture-of-Experts with Subband-Adaptive Routing for Blind All-in-One Image Restoration
Computer Vision and Pattern Recognition
Summary
Fixing images that have different kinds of damage is hard without knowing what happened to them. The authors present WaFIR, a method that breaks down image features into smaller pieces based on frequency, then decides how each piece should be fixed using special experts. This approach works better than earlier methods for tasks like removing noise, rain, or fog from images, even when the problem is mixed or unseen. Their results suggest that looking at image parts by their frequency helps computers fix pictures more accurately.
What this means in practice
- •For photo editing software developers: Add adaptive frequency-based correction to improve automated cleaning of noisy, rainy, or foggy images without user input on degradation type.$Commercial implications: Enables new features in photo editors that automatically repair diverse image defects, enhancing user experience and reducing manual adjustments.
- •For security camera system engineers: Improve image clarity from diverse environmental degradations by integrating a subband-adaptive restoration module to maintain better visual quality.
Authors
Trong Nhan Nguyen, Vu Hoang Minh Doan, Thanh Dat Le, Truong Tien Vo, Thi Thu Ha Vu, Tan Hung Vo, Jaeyeop Choi, Junghwan Oh, Truong-Son Hy
Abstract
Blind all-in-one image restoration must adapt a shared model to heterogeneous and sometimes mixed degradations without a task label at test time. We introduce WaFIR, a decoder module that decomposes features into Haar subbands, summarizes each subband with a compact Fourier descriptor, and routes it to prompt-conditioned spectral modulation experts. Unlike image-level frequency conditioning, WaFIR makes routing decisions at the subband level, enabling frequency-specific processing while retaining the task context learned by the backbone. Under a unified training and evaluation protocol for denoising, deraining, and dehazing, WaFIR improves average PSNR over the three tasks by 0.48 dB over the PromptIR baseline and by 0.38 dB over the recent Up-Restorer. Matched-capacity and routing controls show that the gain is not explained by additional parameters alone. On unseen noise intensities, descriptor-based routing is more robust than a discrete label-conditioned alternative and remains on par under compound degradations, while routing analyses show stable subband-dependent expert use across degradation types. These results support subband frequency structure as an interpretable complement to prompt-based conditioning for blind image restoration. Code is available at https://anonymous.4open.science/r/WaFIR/README.md