Reproduce “Neural Field Representations of Mobile Computational Photography” 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
Ilya Chugunov
The dissertation presents task-designed neural fields fitted directly to phone bursts, using small viewpoint and image changes to estimate depth, separate scene layers, and reconstruct interactive panoramas. Its depth method beats SfSM on Dragon scanned-geometry errors, and layer separation scores 41.80 dB PSNR and 0.986 SSIM on synthetic Geese. Panorama comparisons are limited to scenes with valid COLMAP poses.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.