From Iterative Denoising to Conditional Generation: A Survey on Diffusion Model-Driven Methods for Blind Image Super-Resolution
Abstract
Blind image super-resolution remains a significant challenge in recovering high-quality images from low-resolution inputs with unknown degradations. This paper systematically reviews diffusion models applied to this problem, leveraging their strong generative priors. We categorize methods into three pathways: iterative denoising and posterior sampling approaches (e.g., DDRM, DPS) that condition models during inference; methods utilizing pre-trained diffusion models as universal priors for zero-shot or tuned restoration (e.g., SDUP, DiffBIR); and end-to-end trained conditional diffusion models (e.g., SR3, ResShift). We detail the principles, innovations, and performance of each paradigm and discuss key challenges, including fidelity-perception trade-offs, computational efficiency, and generalization to real-world degradations. This survey provides a structured reference for future diffusion-based blind image restoration research.
Read the paper
Related studies
- Learning from First-Generation Qualitative Approaches in the IS Discipline: An Evolutionary View and Some Implications for Authors and Evaluators (PART 1/2)
- Commentary on the clinical and preclinical dosage limits of interstitially administered magnetic fluids for therapeutic hyperthermia based on current practice and efficacy models
- Prosthesis: New Technological Opportunities and Innovative Biomedical Devices
- The history and future of neural modeling for cochlear implants
- Ryanodine receptors: waking up from refractoriness
- Novel Transgenic Rabbit Model Sheds Light on the Puzzling Role of Matrix Metalloproteinase-12 in Atherosclerosis
Go deeper in Evidence Hub
Evidence Hub adds a study data sheet (design, sample size, evidence level, key numbers), an AI summary and a chat grounded in the paper's text.