When Fast Fourier Transform Meets Transformer for Image Restoration (2024)
Researchers have developed SFHformer, a novel image restoration framework that integrates the Fast Fourier Transform (FFT) with Transformer architecture. This approach leverages both spatial and frequency domains to model local and global image features, addressing challenges across various degradation phenomena. The framework has demonstrated state-of-the-art performance on numerous datasets for tasks like deraining, dehazing, deblurring, and low-light enhancement, offering a favorable balance of performance, parameter size, and computational cost. AI
IMPACT Introduces a novel framework for image restoration that improves performance across multiple degradation tasks.