WEVSR: Adapting Video Diffusion Generators to Real-World Video Super‑Resolution with Wavelet-Enhanced VAE Encoder
Yuying Chen, Liu, Linyan Jiang, Qifan Gao, Xianguo Zhang, Jianhou Gan, Wenqi Ren
Abstract
Recent advances in video diffusion models have demonstrated remarkable generative capability, yet adapting these large pretrained text-to-video (T2V) models to video super‑resolution (VSR) typically encounters challenges, such as artifacts introduced by complex degradations in real-world scenarios and compromised fidelity due to the strong generative capacity of the powerful T2V models. We present WEVSR, a novel approach that adapts a pretrained flow-matching video diffusion transformer to RealVSR. First, we design a task-oriented adaptation strategy that leverages timestep sampling and noise augmentation to enhance detail restoration while preserving structural stability. Second, we propose a lightweight multi-level discrete wavelet transform (DWT) front-end for the VAE encoder, injecting explicit frequency priors into the latent space without modifying the pretrained decoder. Extensive experiments across multiple RealVSR benchmarks demonstrate that WEVSR achieves state-of-the-art performance against existing approaches.
BibTeX
@inproceedings{
chen2026wevsr,
title={{WEVSR}: Video Diffusion Generators for Real-World Video Super\nobreakdash-Resolution with Wavelet-Enhanced {VAE} Encoder},
author={Yuying Chen and Zhirui Liu and Linyan Jiang and Qifan Gao and Xianguo Zhang and Jianhou Gan and Wenqi Ren},
booktitle={Forty-third International Conference on Machine Learning},
year={2026},
url={https://openreview.net/forum?id=WYBM6ES3iH}
}