Reproduce “Edge-Efficient Joint RAW Image Compression and Restoration for Smart Wearable Devices” 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

Yunong Sun, Sicheng Chen, Fei Xia

The authors present a wearable system that compresses noisy RAW camera data for clean backend reconstruction. A lightweight glasses encoder sends a compact bitstream, while a larger decoder jointly denoises and demosaics the image. On SIDD, Ours-M reaches 52.90 dB at 0.2509 bpp, versus 51.34 dB for JXL+NAFNet at nearly the same rate; learned transforms saturate on texture-rich LSDIR at high rates.

Summary by alphaXiv.

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