QUIET-SR: Quantum Image Enhancement Transformer for Single Image Super-Resolution

2025-03-11

AI Summary

To address the dual bottlenecks of excessive parameter counts in classical deep models for single-image super-resolution (SISR) and poor scalability of quantum image processing, this paper proposes the first lightweight super-resolution framework integrating Transformers with variational quantum neural networks (VQNNs). We innovatively design a shift-quantum-window attention mechanism, embedding VQNNs into the Swin Transformer backbone to enable scalable, quantum-enhanced feature extraction on near-term intermediate-scale quantum (NISQ) hardware. A classical–quantum co-training strategy and a hybrid architecture—comprising Swin Transformer, VQNN, and quantized window attention—are employed. Evaluated on MNIST (30.24 PSNR), FashionMNIST (29.76 PSNR), and MedMNIST, our method achieves state-of-the-art performance while significantly reducing model parameters, thereby balancing reconstruction accuracy, computational efficiency, and hardware feasibility.

Technology Category

• Machine Learning: Quantum Machine Learning
• Computer Vision: Multi-modal Vision
• Search and Optimization: Learning to Search

Application Category

• Search and Retrieval-Augmented AI: Efficiency and scalability of Web search engines
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Recent advancements in Single-Image Super-Resolution (SISR) using deep learning have significantly improved image restoration quality. However, the high computational cost of processing high-resolution images due to the large number of parameters in classical models, along with the scalability challenges of quantum algorithms for image processing, remains a major obstacle. In this paper, we propose the Quantum Image Enhancement Transformer for Super-Resolution (QUIET-SR), a hybrid framework that extends the Swin transformer architecture with a novel shifted quantum window attention mechanism, built upon variational quantum neural networks. QUIET-SR effectively captures complex residual mappings between low-resolution and high-resolution images, leveraging quantum attention mechanisms to enhance feature extraction and image restoration while requiring a minimal number of qubits, making it suitable for the Noisy Intermediate-Scale Quantum (NISQ) era. We evaluate our framework in MNIST (30.24 PSNR, 0.989 SSIM), FashionMNIST (29.76 PSNR, 0.976 SSIM) and the MedMNIST dataset collection, demonstrating that QUIET-SR achieves PSNR and SSIM scores comparable to state-of-the-art methods while using fewer parameters. These findings highlight the potential of scalable variational quantum machine learning models for SISR, marking a step toward practical quantum-enhanced image super-resolution.

Problem

• High computational cost in classical SISR models.
• Scalability challenges of quantum image processing.
• Need for efficient quantum-enhanced image super-resolution.

Innovation

• Hybrid Swin transformer with quantum attention
• Minimal qubits for NISQ-era compatibility
• Scalable variational quantum machine learning