Skip to main content
GameDev.net gamedev.net

Super Resolution News

The latest Super Resolution coverage curated for game developers.

Human Vision Constrained Super-Resolution

A new super-resolution approach uses a human-vision model to skip detail work the player likely cannot perceive, cutting compute by 2x or more without hurting perceived quality. The framework adapts …

Research arXiv cs.GR · 2 months, 1 week ago
Learning to Adaptively Allocate Gaussians for Arbitrary-Scale Image Super-Resolution

A new ASR approach, QuADA-GS, tries to make arbitrary-scale upscaling practical by allocating compute only where image detail actually needs it. Instead of treating every pixel uniformly, it routes a …

Research arXiv cs.GR · 2 months, 3 weeks ago
Resonant Brane Splatting for Arbitrary-Scale Super-Resolution

A new super-resolution paper swaps smooth Gaussian splats for richer “Brane” primitives that can carry internal high-frequency detail. The practical goal is fewer overlapping splats, less rasterization pain, and better …

Research arXiv cs.GR · 2 months, 3 weeks ago
GB-LSR: A Fast Local Spectral Image Representation with a Single Global Bandwidth for Continuous Reconstruction and Super-Resolution

A new image representation claims a big speed/quality win for continuous reconstruction: GB-LSR uses one global trainable bandwidth across all patches and still beats matched-budget LIIF/LTE/WIRE baselines by 2.8-3.6 dB …

Research arXiv cs.GR · 3 months, 1 week ago
Multiscale Super Resolution without Image Priors

A new approach to super-resolution imaging could significantly enhance graphics programming. By leveraging low-resolution images at varying scales, developers can achieve clearer, more detailed visuals without relying on traditional image …

Research arXiv cs.GR · 5 months ago
CV-HoloSR: Hologram to hologram super-resolution through volume-upsampling three-dimensional scenes

CV-HoloSR introduces a groundbreaking approach to hologram super-resolution, enhancing volumetric spatial up-sampling while maintaining depth accuracy. This innovation is particularly relevant for graphics programmers and artists, as it significantly improves …

Research arXiv cs.GR · 5 months, 1 week ago
RealOSR: Latent Guidance Boosts Diffusion-based Real-world Omnidirectional Image Super-Resolutions

RealOSR introduces a significant leap in omnidirectional image super-resolution, achieving over 200 times faster inference. This advancement is particularly relevant for graphics programmers and artists, as it enhances the visual …

Research arXiv cs.GR · 6 months, 3 weeks ago
Synthetic Abundance Maps for Unsupervised Super-Resolution of Hyperspectral Remote Sensing Images

A novel unsupervised training framework for hyperspectral image super-resolution could revolutionize how graphics programmers enhance image quality. By using synthetic abundance maps, this method eliminates the need for high-resolution ground …

Research arXiv cs.GR · 7 months, 3 weeks ago
Unsupervised Super-Resolution of Hyperspectral Remote Sensing Images Using Fully Synthetic Training

A new unsupervised training strategy for hyperspectral image super-resolution could revolutionize how developers handle remote sensing data. By utilizing synthetic abundance data, this method eliminates the need for ground truth …

Research arXiv cs.GR · 8 months ago
SplatSuRe: Selective Super-Resolution for Multi-view Consistent 3D Gaussian Splatting

SplatSuRe introduces a novel approach to super-resolution in 3D Gaussian Splatting, selectively enhancing low-resolution views to improve multi-view consistency. This method is particularly beneficial for graphics programmers, as it yields …

Research arXiv cs.GR · 9 months, 3 weeks ago