Reproduce “Deep Networks for Image Super-Resolution with Sparse Prior” 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
Zhaowen Wang, Ding Liu, Jianchao Yang, Wei Han, Thomas Huang
The authors build a single-image super-resolution network that embeds a sparse-coding prior in end-to-end learning. The network uses learned iterative sparse-code updates to reconstruct image patches, and cascades stages for larger upscaling. On Set5, CSCN-MV reaches mean PSNR of 33.26 dB at three-times upscaling, versus 32.75 dB for CNN-L. The result uses bicubic-downsampled tests and multiview processing.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.