Structured pattern embedding and efficient global dependency modeling for lightweight image super-resolution

doi:10.1016/j.cviu.2026.104922

Structured pattern embedding and efficient global dependency modeling for lightweight image super-resolution

2026-08-25

0

PRE

AI

DOI：10.1016/j.cviu.2026.104922

Original

Original request for help

Save

Abstract

En 中文

• Achieves state-of-the-art performance under strict parameter and FLOP budgets. • Employs parallel-hierarchical pattern embedding for rich, diverse local features. • Bridges local fine-grained detail and global context in lightweight SISR.

Keywords:

Lightweight image super-resolution

Efficient global modeling

Pattern selection

Spatial pattern embedding

Structured pattern embedding and efficient global dependency modeling for lightweight image super-resolution

Abstract

Journal

Organization

Cited Papers