DUO-VSR: Dual-Stream Distillation for One-Step Video Super-Resolution
2026-03-23 • Computer Vision and Pattern Recognition
Computer Vision and Pattern Recognition
AI summaryⓘ
The authors address the problem that making video super-resolution using diffusion models is usually slow and unstable. They propose DUO-VSR, a three-step method that combines two techniques: distribution matching distillation for faster sampling and adversarial learning for better image quality. Their approach first stabilizes training, then trains joint models to improve realism, and finally refines results based on human-like preferences. Experiments show their method improves speed and visual quality compared to earlier one-step methods.
video super-resolutiondiffusion modelsdistribution matching distillationadversarial supervisionGANone-step generationtraining stabilityperceptual quality
Authors
Zhengyao Lv, Menghan Xia, Xintao Wang, Kwan-Yee K. Wong
Abstract
Diffusion-based video super-resolution (VSR) has recently achieved remarkable fidelity but still suffers from prohibitive sampling costs. While distribution matching distillation (DMD) can accelerate diffusion models toward one-step generation, directly applying it to VSR often results in training instability alongside degraded and insufficient supervision. To address these issues, we propose DUO-VSR, a three-stage framework built upon a Dual-Stream Distillation strategy that unifies distribution matching and adversarial supervision for one-step VSR. Firstly, a Progressive Guided Distillation Initialization is employed to stabilize subsequent training through trajectory-preserving distillation. Next, the Dual-Stream Distillation jointly optimizes the DMD and Real-Fake Score Feature GAN (RFS-GAN) streams, with the latter providing complementary adversarial supervision leveraging discriminative features from both real and fake score models. Finally, a Preference-Guided Refinement stage further aligns the student with perceptual quality preferences. Extensive experiments demonstrate that DUO-VSR achieves superior visual quality and efficiency over previous one-step VSR approaches.