Reproduce “D-TensoRF: Tensorial Radiance Fields for Dynamic Scenes” 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
Hankyu Jang, Daeyoung Kim
D-TensoRF reconstructs dynamic scenes by adding time as a dimension of TensoRF’s factorized grid and smoothing features across nearby times, rather than learning a separate deformation network. On eight scenes, CP768 achieves LPIPS 0.03 with a 1.8 MB model, but lower PSNR/SSIM and longer training than TiNeuVox-B. It can smudge with large adjacent-time motion.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.