R2GDN: RepGhost based residual dense network for image super-resolution

Tianyu Li, Xiaoshi Jin, Qiang Liu, Xi Liu, Zehang Yuan, Tianyang Liang, Jia Lou, Yangfan Rao · 2025

Read R2GDN: RepGhost based residual dense network for image super-resolution by Tianyu Li, Xiaoshi Jin, Qiang Liu, Xi Liu, Zehang Yuan, Tianyang Liang, Jia Lou, Yangfan Rao (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/pmc-12700410.txt.

Description

Tianyu Li leads a team of eight authors in 2025 in designing a lighter network for image super-resolution, the task of rebuilding a sharp picture from a small one. Existing networks do this well but demand heavy computing and memory, which keeps them off phones and other edge devices.

Their network, R2GDN, reuses features across dense residual blocks and borrows a reparameterization layer based on the lean RepGhost design. Cheap operations squeeze extra features out of the ones already computed, and the training-time structure folds away for fast running. The result targets clear upscaling that modest hardware can actually run.

Catalog record

CREATORTianyu Li, Xiaoshi Jin, Qiang Liu, Xi Liu, Zehang Yuan, Tianyang Liang, Jia Lou, Yangfan Rao
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE54 KB
VERSION1.0
LICENSECC BY 4.0
RIGHTSCC BY
SOURCEEurope PMC

Source: https://europepmc.org/articles/PMC12700410

Browse the full library · About the Tower of Babel