ResQu - Quaternion Wavelet-Conditioned Diffusion for Super-Resolution

Overview

ResQu is a Quaternion Wavelet-Conditioned Diffusion Model designed to enhance image super-resolution tasks. This method introduces quaternion wavelet embeddings to improve feature representation, enabling high-fidelity reconstructions with improved perceptual quality.

Key Features

• Wavelet-Based Preprocessing: Integrates quaternion wavelet decomposition to enhance texture details.
• Diffusion Model for Super-Resolution: Uses a conditional denoising diffusion model (StableSR baseline).
• State-of-the-Art Performance: Achieves +15% PSNR improvement over traditional super-resolution models.
• Multi-Scale Feature Conditioning: Enables enhanced frequency-aware feature learning.

Paper

• Title: Quaternion Wavelet-Conditioned Diffusion for Image Super-Resolution
• Status: Under Review at IJCNN 2025

Methodology

ResQu leverages a multi-scale frequency decomposition using quaternion wavelets, feeding wavelet-conditioned embeddings into a diffusion model. The architecture consists of:

1. Wavelet Decomposition: Extracts low- and high-frequency components.
2. Quaternion Embeddings: Encodes spatial-frequency information.
3. Diffusion-Based Enhancement: Refines reconstructions through iterative denoising.

StableSR

@article{wang2024exploiting,
  author = {Wang, Jianyi and Yue, Zongsheng and Zhou, Shangchen and Chan, Kelvin C.K. and Loy, Chen Change},
  title = {Exploiting Diffusion Prior for Real-World Image Super-Resolution},
  journal = {International Journal of Computer Vision},
  year = {2024}
}

Quave

@misc{sigillo2024generalizing,
      title={Generalizing Medical Image Representations via Quaternion Wavelet Networks}, 
      author={Luigi Sigillo and Eleonora Grassucci and Aurelio Uncini and Danilo Comminiello},
      year={2024},
      eprint={2310.10224},
      archivePrefix={arXiv},
      primaryClass={eess.IV}
}

---

Acknowledgment

This project is based on the StableSR framework developed by researchers at Nanyang Technological University and the QUAVE framework developed at Sapienza University. Their combined capabilities offer a powerful approach to image super-resolution.