AI Paper Atlas

Research map / 3D vision / NeRF and radiance fields

Tetra-NeRF: Representing Neural Radiance Fields Using Tetrahedra

Jonas Kulhanek, Torsten Sattler

ICCV 2023 · 3D vision › NeRF and radiance fields · approach: Dynamic and editable NeRF · 58 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 a continuous tetrahedra field instead of a discrete point cloud when Point-NeRF growing and pruning is disabled (ablation)Tetra-NeRF vs Point-NeRF · Blender shipPSNR: 19.35 -> 31.13Improved
use adaptive tetrahedra instead of dense voxel grid at similar parameter count (ablation)Tetra-NeRF · Blender legoPSNR: 18.81 -> 33.79Improved
use same COLMAP points as Point-NeRF with tetrahedra field instead of point fieldTetra-NeRF · BlenderPSNR: 31.77 -> 32.53Improved
Tetra-NeRF compared with mip-NeRF and instant-NGP on Blender PSNRTetra-NeRF · BlenderPSNR: 34.51 -> 32.53 (mip-NeRF vs Tetra-NeRF)Hurt
use sparse instead of dense point cloud on synthetic Blender scenes (ablation)Tetra-NeRF · Blender legoPSNR: 33.79 -> 29.77Hurt
use sparse instead of dense point cloud on real-world scenes (ablation)Tetra-NeRF · Mip-NeRF 360 bonsaiPSNR: 28.34 -> 28.12No clear effect
increase number of input points (ablation)Tetra-NeRF · Tanks and Temples and Mip-NeRF 360 scenesreconstruction qualityImproved
Tetra-NeRF versus Point-NeRF on Tanks and TemplesTetra-NeRF · Tanks and TemplesLPIPS: 0.090 -> 0.059Improved
Tetra-NeRF versus mip-NeRF 360 on SSIMTetra-NeRF · Mip-NeRF 360 (mean)SSIM: 0.786 -> 0.717Hurt
cap number of intersected triangles per ray at 512Tetra-NeRF · large scenesreconstruction qualityHurt

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 0 of 2 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