AI Paper Atlas

Research map / 3D vision / 3D generation

Geometry-Free View Synthesis: Transformers and No 3D Priors

Robin Rombach, Patrick Esser, Björn Ommer

ICCV 2021 · 3D vision › 3D generation · approach: Learning 3D reconstruction from images · 75 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (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
add explicit warping to the implicit model (hybrid) (ablation)impl.-depth · ACIDNLL: 5.353 -> 5.341No clear effect
add monocular depth input to the implicit model (ablation)impl.-nodepth · RealEstate10KPSNR: 12.16 -> 12.66Improved
add depth as extra tokens (catdepth) instead of merged codes (ablation)impl.-depth · RealEstate10KNLL: 4.836 -> 4.860No clear effect
warp learnable embeddings instead of fixed VQGAN features (ablation)expl.-feat · ACIDNLL: 5.657 -> 5.416Improved
replace warped features with implicit conditioning on depth and source (ablation)expl.-emb · ACIDNLL: 5.416 -> 5.353Improved
use a probabilistic transformer instead of a deterministic explicit model (ablation)expl.-det · RealEstate10KFID: 66.66 -> 49.15Improved
take 32 samples and keep the best instead of one sampleimpl.-depth · RealEstate10KPSNR: 12.66 -> 14.47Improved
take 32 samples and keep the best instead of one sampleimpl.-depth · ACIDPSNR: 15.53 -> 16.80Improved
use a larger VQGAN latent code (f8 instead of f16) (ablation)VQGAN · reconstructionPSNR: 21.14 -> 25.23Improved
evaluate all test pairs instead of the better of two viewsSynSin · RealEstate10KPSNR: 23.03 -> 19.75Hurt
add 50% larger viewpoint changes to the evaluationSynSin · RealEstate10KPSNR: 23.03 -> 16.05Hurt
add the predicted depth only via the implicit variant without any depth, versus explicit warped images (ablation)impl.-nodepth · RealEstate10KNLL: 4.924 -> 4.956Hurt

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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