AI Paper Atlas

Research map / 3D vision / 3D generation

UpFusion: Novel View Diffusion from Unposed Sparse View Observations

Bharath Raj Nagoor Kani, Hsin-Ying Lee, Sergey Tulyakov, Shubham Tulsiani

ECCV 2024 · 3D vision › 3D generation · approach: Diffusion-based novel view synthesis · 9 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
use pose-free UpFusion instead of SparseFusion with RelPose++ estimated posesUpFusion (3D) · Co3Dv2 seen, 6 viewsPSNR-A: 17.12 -> 18.96Improved
add more unposed input views (1 to 6)UpFusion (3D) · Co3Dv2 seenPSNR-A: 18.17 -> 18.96Improved
add more views to SparseFusion with RelPose++ posesSparseFusion (RelPose++) · Co3Dv2 seenPSNR-A: 17.76 -> 17.12 (3V to 6V)Hurt
distill a 3D-consistent NeRF via score distillation from the 2D diffusion modelUpFusion (2D) · Co3Dv2 seen, 3 viewsPSNR-A: 17.12 -> 18.68Improved
use diffusion (UpFusion 2D) instead of regression (UpSRT)UpSRT · Co3Dv2 seen, 3 viewsLPIPS-A: 0.32 -> 0.22Improved
use diffusion (UpFusion 2D) instead of regression (UpSRT)UpSRT · Co3Dv2 seen, 3 viewsPSNR-A: 17.75 -> 17.12Hurt
add set-latent attention conditioning to decoder features (ablation)UpFusion 2D (teddybear) · Co3Dv2 teddybear, 3 viewsPSNR-A: 15.28 -> 16.11Improved
add decoder features to set-latent conditioning (ablation)UpFusion 2D (teddybear) · Co3Dv2 teddybear, 3 viewsPSNR-A: 13.38 -> 16.11Improved
evaluate on unseen categoriesUpFusion (3D) · Co3Dv2 unseen vs seen, 3 viewsPSNR-A: 18.68 (seen) vs 18.83 (unseen)No clear effect
use six unposed views instead of pose-optimizing FORGEUpFusion-dagger (3D) · GSO, 6 viewsPSNR: 17.40 -> 22.51Improved
use UpFusion-dagger with a single view instead of Zero-1-to-3 (threestudio)single-view 3D · GSO, 1 viewPSNR: 21.71 -> 20.52Hurt

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