AI Paper Atlas

Research map / 3D vision / Face and shape

TAPS3D: Text-Guided 3D Textured Shape Generation From Pseudo Supervision

Jiacheng Wei, Hao Wang, Jiashi Feng, Guosheng Lin, Kim-Hui Yap

CVPR 2023 · 3D vision › Face and shape · approach: 3D shape generative modeling · 21 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 TAPS3D instead of CLIP-NeRFText-guided 3D generation · CarFID: 67.8 -> 20.1Improved
Use TAPS3D instead of CLIP-NeRFText-guided 3D generation · ChairFID: 48.4 -> 43.7Improved
Use TAPS3D instead of TITG3SGText-guided 3D generation · ChairCLIP R-Precision: 9.05 ± 1.92 -> 20.19 ± 1.89Improved
Use TAPS3D instead of TITG3SGText-guided 3D generation · TableCLIP R-Precision: 6.67 ± 3.25 -> 15.21 ± 3.15Improved
Use TAPS3D instead of CLIP-Forge3D geometry generation · ChairFPD: 825.96 -> 342.23Improved
Use TAPS3D instead of TITG3SG3D geometry generation · TableFPD: 1639.68 -> 1468.43Improved
Use pseudo-text CLIP features rather than image CLIP features as training input (ablation)TAPS3D · ShapeNet Car, Chair, Table, MotorbikeFID: Car: 68.7 -> 21.7; Chair: 81.1 -> 44.8; Table: 94.3 -> 43.2; Motorbike: 146.5 -> 57.2Improved
Add CLIP loss (ablation)TAPS3D · ShapeNet CarFID: 83.7 -> 21.7Improved
Add CLIP loss (ablation)TAPS3D · ShapeNet ChairFID: 148.9 -> 44.8Improved
Add image regularization loss (ablation)TAPS3D · ShapeNet CarFID: 59.8 -> 21.7Improved
Add image regularization loss (ablation)TAPS3D · ShapeNet ChairFID: 144.2 -> 44.8Improved
Add background augmentation (ablation)TAPS3D · ShapeNet CarFID: 55.1 -> 21.7Improved

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