Resonant Brane Splatting augments Gaussians with internal Hermite modes to create Branes that reconstruct fine details with fewer primitives, plus an efficient differentiable rasterizer for arbitrary-scale super-resolution.
Pixel to gaussian: Ultra-fast continuous super-resolution with 2d gaussian modeling
8 Pith papers cite this work. Polarity classification is still indexing.
8
Pith papers citing it
abstract
Arbitrary-scale super-resolution (ASSR) aims to reconstruct high-resolution (HR) images from low-resolution (LR) inputs with arbitrary upsampling factors using a single model, addressing the limitations of traditional SR methods constrained to fixed-scale factors (\textit{e.g.}, $\times$ 2). Recent advances leveraging implicit neural representation (INR) have achieved great progress by modeling coordinate-to-pixel mappings. However, the efficiency of these methods may suffer from repeated upsampling and decoding, while their reconstruction fidelity and quality are constrained by the intrinsic representational limitations of coordinate-based functions. To address these challenges, we propose a novel ContinuousSR framework with a Pixel-to-Gaussian paradigm, which explicitly reconstructs 2D continuous HR signals from LR images using Gaussian Splatting. This approach eliminates the need for time-consuming upsampling and decoding, enabling extremely fast arbitrary-scale super-resolution. Once the Gaussian field is built in a single pass, ContinuousSR can perform arbitrary-scale rendering in just 1ms per scale. Our method introduces several key innovations. Through statistical ana
citation-role summary
method 1
citation-polarity summary
fields
cs.CV 8roles
method 1polarities
use method 1representative citing papers
EIC-LIE uses an event-illumination collaborative module and illumination-aware event filter plus a new real-world dataset to improve low-light image enhancement over prior methods.
GS-STVSR achieves state-of-the-art continuous spatio-temporal video super-resolution quality with nearly constant inference time at standard scales and over 3x speedup at extreme scales using 2D Gaussian Splatting.
A 2D Gaussian Splatting framework with MRI-specific anatomical priors, physics-constrained intensity modeling, and meta-learning domain adaptation achieves state-of-the-art MRI super-resolution while showing that intermediate input resolutions can outperform maximum-resolution inputs.
A single reference-guided flow-matching model synthesizes continuous-scale (3–20×) high-detail garment images from unaligned full-view and close-up product photos, matching per-instance fine-tuning quality at far lower cost.
GSPan introduces continuous 2D Gaussian primitives for band-wise residuals in pansharpening to support arbitrary-scale fusion without retraining.
GaussianHSI uses Voronoi-guided bilateral 2D Gaussian splatting plus a spectral detail enhancement module to perform arbitrary-scale hyperspectral image super-resolution.
A new dataset with high-fidelity close-up garment images and full/close-up try-on videos plus the VGID metric enables better texture and structure preservation in high-resolution video virtual try-on.
citing papers explorer
-
Resonant Brane Splatting for Arbitrary-Scale Super-Resolution
Resonant Brane Splatting augments Gaussians with internal Hermite modes to create Branes that reconstruct fine details with fewer primitives, plus an efficient differentiable rasterizer for arbitrary-scale super-resolution.
-
Event-Illumination Collaborative Low-light Image Enhancement with a High-resolution Real-world Dataset
EIC-LIE uses an event-illumination collaborative module and illumination-aware event filter plus a new real-world dataset to improve low-light image enhancement over prior methods.
-
GS-STVSR: Ultra-Efficient Continuous Spatio-Temporal Video Super-Resolution via 2D Gaussian Splatting
GS-STVSR achieves state-of-the-art continuous spatio-temporal video super-resolution quality with nearly constant inference time at standard scales and over 3x speedup at extreme scales using 2D Gaussian Splatting.
-
PhyMRI-SR: Toward Physics-Aware MRI Image Super-Resolution
A 2D Gaussian Splatting framework with MRI-specific anatomical priors, physics-constrained intensity modeling, and meta-learning domain adaptation achieves state-of-the-art MRI super-resolution while showing that intermediate input resolutions can outperform maximum-resolution inputs.
-
GarmentZoom: Generating Zoomable Images from Garment Listings
A single reference-guided flow-matching model synthesizes continuous-scale (3–20×) high-detail garment images from unaligned full-view and close-up product photos, matching per-instance fine-tuning quality at far lower cost.
-
GSPan: A Continuous Gaussian Primitive Representation for Arbitrary-Scale Pansharpening
GSPan introduces continuous 2D Gaussian primitives for band-wise residuals in pansharpening to support arbitrary-scale fusion without retraining.
-
Voronoi-guided Bilateral 2D Gaussian Splatting for Arbitrary-Scale Hyperspectral Image Super-Resolution
GaussianHSI uses Voronoi-guided bilateral 2D Gaussian splatting plus a spectral detail enhancement module to perform arbitrary-scale hyperspectral image super-resolution.
-
Eevee: Towards Close-up High-resolution Video-based Virtual Try-on
A new dataset with high-fidelity close-up garment images and full/close-up try-on videos plus the VGID metric enables better texture and structure preservation in high-resolution video virtual try-on.