Wavelet-Space Representations for Neural Super-Resolution in Rendering Pipelines

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Poudel, Prateek, Aryal, Prashant, Kunwar, Kirtan, Nepal, Navin, Kshatri, Dinesh Baniya
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866909798207848448
author Poudel, Prateek
Aryal, Prashant
Kunwar, Kirtan
Nepal, Navin
Kshatri, Dinesh Baniya
author_facet Poudel, Prateek
Aryal, Prashant
Kunwar, Kirtan
Nepal, Navin
Kshatri, Dinesh Baniya
contents We investigate the use of wavelet-space feature decomposition in neural super-resolution for rendering pipelines. Building on recent neural upscaling frameworks, we introduce a formulation that predicts stationary wavelet coefficients rather than directly regressing RGB values. This frequency-aware decomposition separates low- and high-frequency components, enabling sharper texture recovery and reducing blur in challenging regions. Unlike conventional wavelet transforms, our use of the stationary wavelet transform (SWT) preserves spatial alignment across subbands, allowing the network to integrate G-buffer attributes and temporally warped history frames in a shift-invariant manner. The predicted coefficients are recombined through inverse wavelet synthesis, producing resolution-consistent reconstructions across arbitrary scale factors. We conduct extensive evaluations and ablations, showing that incorporating SWT improves both fidelity and perceptual quality with only modest overhead, while remaining compatible with standard rendering architectures. Taken together, our results suggest that wavelet-domain neural super-resolution provides a principled and efficient path toward higher-quality real-time rendering, with broader implications for neural rendering and graphics applications.
format Preprint
id arxiv_https___arxiv_org_abs_2508_16024
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Wavelet-Space Representations for Neural Super-Resolution in Rendering Pipelines
Poudel, Prateek
Aryal, Prashant
Kunwar, Kirtan
Nepal, Navin
Kshatri, Dinesh Baniya
Graphics
Computer Vision and Pattern Recognition
We investigate the use of wavelet-space feature decomposition in neural super-resolution for rendering pipelines. Building on recent neural upscaling frameworks, we introduce a formulation that predicts stationary wavelet coefficients rather than directly regressing RGB values. This frequency-aware decomposition separates low- and high-frequency components, enabling sharper texture recovery and reducing blur in challenging regions. Unlike conventional wavelet transforms, our use of the stationary wavelet transform (SWT) preserves spatial alignment across subbands, allowing the network to integrate G-buffer attributes and temporally warped history frames in a shift-invariant manner. The predicted coefficients are recombined through inverse wavelet synthesis, producing resolution-consistent reconstructions across arbitrary scale factors. We conduct extensive evaluations and ablations, showing that incorporating SWT improves both fidelity and perceptual quality with only modest overhead, while remaining compatible with standard rendering architectures. Taken together, our results suggest that wavelet-domain neural super-resolution provides a principled and efficient path toward higher-quality real-time rendering, with broader implications for neural rendering and graphics applications.
title Wavelet-Space Representations for Neural Super-Resolution in Rendering Pipelines
topic Graphics
Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2508.16024