Search papers, labs, and topics across Lattice.
This paper introduces a perceptually regularized diffusion model for image super-resolution that enhances the recovery of high-resolution images from low-resolution inputs by incorporating perceptual-loss-based regularization. By addressing the limitations of traditional pixel-domain noise-prediction loss, the proposed framework improves training convergence and preserves fine image structures, which are often lost in standard approaches. Experimental results on benchmark datasets show that this method achieves superior perceptual quality while maintaining competitive distortion metrics, underscoring its effectiveness in practical applications.
Incorporating perceptual regularization into diffusion models significantly enhances image super-resolution, preserving fine details that traditional methods often overlook.
Image super-resolution, which aims to reconstruct high-resolution images from their low-resolution observations, is fundamental to medical imaging, remote sensing, surveillance, microscopy, and scientific visualization. Traditional model-based methods formulate super-resolution as an inverse problem with hand-crafted regularization priors. While interpretable and theoretically grounded, they rely on fixed assumptions and require computationally intensive iterative solvers. Deep learning methods offer data-driven flexibility by learning nonlinear mappings from low- to high-resolution images, among which diffusion models have achieved particularly impressive perceptual quality. However, the standard diffusion training objective is a pixel-domain noise-prediction loss that does not explicitly enforce perceptual fidelity, which can lead to oversmoothing and loss of fine image structure. To address these limitations, we propose a perceptually regularized diffusion framework that incorporates prior knowledge through perceptual-loss-based regularization, improving training convergence and encouraging the recovery of meaningful image features. Experiments on benchmark datasets demonstrate improved perceptual quality and competitive distortion metrics, highlighting the effectiveness of regularization for diffusion-based super resolution.
