VINS-120K supplies the first large-scale set of instruction-image-edited-image triplets at ultra-high resolution together with an adaptation strategy that improves detail synthesis.
hub
Real-esrgan: Training real-world blind super-resolution with pure synthetic data
11 Pith papers cite this work. Polarity classification is still indexing.
11
Pith papers citing it
hub tools
citation-role summary
method 3
citation-polarity summary
fields
cs.CV 11roles
method 3polarities
use method 3representative citing papers
VOSR shows that competitive generative image super-resolution with faithful structures can be achieved by training a diffusion-style model from scratch on visual data alone, using a vision encoder for guidance and a restoration-oriented sampling strategy.
Face2Scene uses facial restoration as an oracle to derive degradation codes that condition a diffusion model for restoring the entire degraded scene.
Semantic pseudo-pairing via DINOv2 embeddings and fused Gromov-Wasserstein optimal transport enables training a 7K-parameter CNN for unpaired smartphone ISP, achieving 22.569 PSNR on the NTIRE 2026 challenge test set.
MagicBokeh uses a single diffusion model with alternative training, focus-aware masked attention, and degradation-aware depth estimation to produce photorealistic bokeh on low-res zoomed images.
SlimDiffSR uses uncertainty-guided timestep assignment and structured pruning with frequency- and direction-separable convolutions plus MMD distillation to create a 200x faster, 20x smaller diffusion SR model for remote sensing while retaining competitive quality.
GaussianZoom enables high-fidelity extreme zoom-in 3D rendering from low-res inputs via an iterative framework combining geometry-consistent modeling, depth-based super-resolution, VLM detail synthesis, and an expandable continuous Level-of-Detail hierarchy.
FRAMER improves real-world super-resolution by decomposing features into low- and high-frequency bands via FFT, applying intra- and inter-contrastive losses with adaptive modulators, and using the final layer as teacher for intermediate layers during diffusion denoising.
BIR-Adapter adds a parameter-efficient attention adapter and guided sampling to pretrained diffusion models, achieving competitive blind image restoration performance with up to 36x fewer trained parameters and enabling extension to new degradation types.
A dual-branch training-free ensemble fuses a hybrid attention network with a Mamba-based model via weighted combination to enhance super-resolution PSNR on DIV2K x4.
Two new lightweight modules for diffusion-based real-world image super-resolution deliver competitive perceptual quality and better structure preservation on DIV2K and RealSR datasets.
citing papers explorer
-
VINS-120K: Ultra High-Resolution Image Editing with A Large-Scale Dataset
VINS-120K supplies the first large-scale set of instruction-image-edited-image triplets at ultra-high resolution together with an adaptation strategy that improves detail synthesis.
-
VOSR: A Vision-Only Generative Model for Image Super-Resolution
VOSR shows that competitive generative image super-resolution with faithful structures can be achieved by training a diffusion-style model from scratch on visual data alone, using a vision encoder for guidance and a restoration-oriented sampling strategy.
-
Face2Scene: Using Facial Degradation as an Oracle for Diffusion-Based Scene Restoration
Face2Scene uses facial restoration as an oracle to derive degradation codes that condition a diffusion model for restoring the entire degraded scene.
-
Lightweight Unpaired Smartphone ISP Transfer with Semantic Pseudo-Pairing
Semantic pseudo-pairing via DINOv2 embeddings and fused Gromov-Wasserstein optimal transport enables training a 7K-parameter CNN for unpaired smartphone ISP, achieving 22.569 PSNR on the NTIRE 2026 challenge test set.
-
Towards Photorealistic and Efficient Bokeh Rendering via Diffusion Framework
MagicBokeh uses a single diffusion model with alternative training, focus-aware masked attention, and degradation-aware depth estimation to produce photorealistic bokeh on low-res zoomed images.
-
SlimDiffSR: Toward Lightweight and Efficient Remote Sensing Image Super-Resolution via Diffusion Model Distillation
SlimDiffSR uses uncertainty-guided timestep assignment and structured pruning with frequency- and direction-separable convolutions plus MMD distillation to create a 200x faster, 20x smaller diffusion SR model for remote sensing while retaining competitive quality.
-
GaussianZoom: Progressive Zoom-in Generative 3D Gaussian Splatting with Geometric and Semantic Guidance
GaussianZoom enables high-fidelity extreme zoom-in 3D rendering from low-res inputs via an iterative framework combining geometry-consistent modeling, depth-based super-resolution, VLM detail synthesis, and an expandable continuous Level-of-Detail hierarchy.
-
FRAMER: Frequency-Aligned Self-Distillation with Adaptive Modulation Leveraging Diffusion Priors for Real-World Image Super-Resolution
FRAMER improves real-world super-resolution by decomposing features into low- and high-frequency bands via FFT, applying intra- and inter-contrastive losses with adaptive modulators, and using the final layer as teacher for intermediate layers during diffusion denoising.
-
BIR-Adapter: A parameter-efficient diffusion adapter for blind image restoration
BIR-Adapter adds a parameter-efficient attention adapter and guided sampling to pretrained diffusion models, achieving competitive blind image restoration performance with up to 36x fewer trained parameters and enabling extension to new degradation types.
-
Training-Free Model Ensemble for Single-Image Super-Resolution via Strong-Branch Compensation
A dual-branch training-free ensemble fuses a hybrid attention network with a Mamba-based model via weighted combination to enhance super-resolution PSNR on DIV2K x4.
-
Degradation-Aware and Structure-Preserving Diffusion for Real-World Image Super-Resolution
Two new lightweight modules for diffusion-based real-world image super-resolution deliver competitive perceptual quality and better structure preservation on DIV2K and RealSR datasets.