Spectral and Trajectory Regularization for Diffusion Transformer Super-Resolution
This work is significant for researchers and practitioners working on efficient high-quality image super-resolution by enabling effective one-step distillation for Diffusion Transformers, addressing specific artifact issues.
This paper addresses the challenge of one-step distillation for Diffusion Transformer (DiT) architectures in real-world image super-resolution (Real-ISR), which often suffer from trajectory mismatch and grid-like artifacts. The proposed StrSR framework, using spectral and trajectory regularization, achieves state-of-the-art performance in Real-ISR, improving both quantitative metrics and visual perception.
Diffusion transformer (DiT) architectures show great potential for real-world image super-resolution (Real-ISR). However, their computationally expensive iterative sampling necessitates one-step distillation. Existing one-step distillation methods struggle with Real-ISR on DiT. They suffer from fundamental trajectory mismatch and generate severe grid-like periodic artifacts. To tackle these challenges, we propose StrSR, a novel one-step adversarial distillation framework featuring spectral and trajectory regularization. Specifically, we propose an asymmetric discriminative distillation architecture to bridge the trajectory gap. Additionally, we design a frequency distribution matching strategy to effectively suppress DiT-specific periodic artifacts caused by high-frequency spectral leakage. Extensive experiments demonstrate that StrSR achieves state-of-the-art performance in Real-ISR, across both quantitative metrics and visual perception. The code and models will be released at https://github.com/jkwang28/StrSR .