Reproduce “Progressive Multi-Scale Super-Resolution via Cascaded Bicubic Anchors and Inter-Scale Mutual Refinement” on OpenResearch

Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.

Universal app for Apple silicon and Intel. Requires macOS 11 or later.

About this paper

Chun-An Lin, Yu-Chia Chang, Chi-Mao Fan, Yen-Chieh Ouyang, Tsung-Jung Liu

The authors present PMS-SR, a super-resolution model that produces ×2, ×4, and ×8 renditions from one low-resolution image in one pass. Each scale builds on a bicubic enlargement of the preceding prediction, while inter-scale mutual refinement lets finer and coarser features inform one another. Under the protocol, it leads compared multi-scale methods at ×4 and ×8, but trails single-scale Transformers at ×2.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.