AI Paper Atlas

Research map / 3D vision / NeRF and radiance fields

Generative Occupancy Fields for 3D Surface-Aware Image Synthesis

NeurIPS 2021 · 3D vision › NeRF and radiance fields · approach: Single-view and streaming radiance fields · 4 citations (OpenAlex)

Read the paperCheck your idea against 108,967 papers

AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
remove normal regularization L_normal (ablation)GOF · BFMweighted depth variance: 2.54 -> 2.66Helped
remove opacity regularization L_opacity (ablation)GOF · CelebAInception Score: 2.87 -> 2.91No clear effect
remove opacity regularization L_opacity (ablation)GOF · Catsweighted depth variance: 4.28 -> 4.42Helped
remove shrinking sampling process (ablation)GOF · BFMnormal map qualityHelped
replace pi-GAN cumulative rendering with GOF occupancy transitionpi-GAN · BFMMAD: 20.46 -> 13.81Helped
replace pi-GAN cumulative rendering with GOF occupancy transitionpi-GAN · BFMSIDE: 0.727 -> 0.779No clear effect
replace pi-GAN cumulative rendering with GOF occupancy transitionpi-GAN · CelebAdepth variance: 14.58 -> 4.58Helped
replace pi-GAN cumulative rendering with GOF occupancy transitionpi-GAN · CelebAmean curvature: 25.94 -> 23.13Helped
render with surface point only instead of multiple pointsGOF · 256x256 CPU renderingtime per image: 78s -> 56sHelped
apply GOF on CARLApi-GAN · CARLAFID: 29.6 -> 29.3No clear effect

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy