Spectral and Trajectory Regularization for Diffusion Transformer Super-Resolution

2026-03-06

AI Summary

This work addresses the challenge of high-frequency leakage and grid-like periodic artifacts in one-step real-image super-resolution using Diffusion Transformers (DiT), which arise from trajectory misalignment during distillation. To mitigate these issues, the authors propose the StrSR framework, which innovatively integrates adversarial distillation, spectral regularization, and trajectory regularization. Specifically, an asymmetric discriminator-based distillation architecture aligns the teacher–student trajectories, while a frequency-domain distribution matching strategy suppresses periodic artifacts. This approach effectively reduces trajectory deviation and spectral distortion inherent in DiT-based one-step super-resolution, achieving state-of-the-art performance in both quantitative metrics and perceptual visual quality on real-image super-resolution benchmarks.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Machine Learning: Semi-Supervised Learning
• Search and Optimization: Learning to Search

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Responsible Web: Data and user privacy-enhancing technologies for the Web
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs

Abstract

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 .

Problem

• Diffusion Transformer
• Real-world Image Super-Resolution
• One-step Distillation
• Trajectory Mismatch
• Periodic Artifacts

Innovation

• Spectral Regularization
• Trajectory Regularization
• Diffusion Transformer
• One-step Distillation
• Real-world Super-Resolution