Research map / 3D vision / 3D generation
Debiasing Diffusion Priors via 3D Attention for Consistent Gaussian Splatting
Shilong Jin, Haoran Duan, Litao Hua, Wentao Huang, Yuan Zhou
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 changed | Where | Measured | Effect |
|---|---|---|---|
| add HAM to baseline (ablation) | Baseline · 3D generation (text-to-3D) | Janus frequency f_mf (%): 35.6 -> 20.0 | Improved |
| add 3D-AAG to baseline (ablation) | Baseline · 3D generation (text-to-3D) | Inconsistency frequency f_inc (%): 62.2 -> 44.4 | Improved |
| add 3D-AAG to HAM (full TD-Attn) (ablation) | TD-Attn · 3D generation (text-to-3D) | Inconsistency frequency f_inc (%): 57.8 -> 37.8 | Improved |
| add HAM to 3D-AAG (full TD-Attn) (ablation) | TD-Attn · 3D generation (text-to-3D) | Janus frequency f_mf (%): 24.4 -> 17.8 | Improved |
| add HAM and 3D-AAG together (ablation) | Baseline · 3D generation (text-to-3D) | Janus frequency f_mf (%): 35.6 -> 17.8 | Improved |
| add HAM and 3D-AAG together (ablation) | Baseline · 3D generation (text-to-3D) | CLIP_sim: 0.307 -> 0.314 | No clear effect |
| add HAM to EditSplat (ablation) | EditSplat · 3D editing (EditSplat) | CLIP_sim: 0.253 -> 0.272 | Improved |
| add 3D-AAG to EditSplat (ablation) | EditSplat · 3D editing (EditSplat) | CLIP_sim: 0.253 -> 0.261 | Improved |
| add HAM and 3D-AAG to EditSplat (ablation) | EditSplat · 3D editing (EditSplat) | User study: 4.18 -> 6.34 | Improved |
| add TD-Attn to DreamScene | DreamScene · 3D generation (text-to-3D) | Inconsistency frequency f_inc (%): 66.7 -> 53.3 | Improved |
| add TD-Attn to LucidDreamer | LucidDreamer · 3D generation (text-to-3D) | ImageReward: -0.386 -> 0.124 | Improved |
| add TD-Attn to GCS-BEG | GCS-BEG · 3D generation (text-to-3D) | Inconsistency frequency f_inc (%): 60.0 -> 26.7 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- LucidDreamer — LucidDreamer: Towards High-Fidelity Text-to-3D Generation via Interval Score Matching (CVPR 2024) · better on 4 of 4 shared results
- DreamScene — DreamScene: 3D Gaussian-based Text-to-3D Scene Generation via Formation Pattern Sampling (ECCV 2024) · better on 4 of 4 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
- SweetDreamer: Aligning Geometric Priors in 2D diffusion for Consistent Text-to-3D (ICLR 2024 · similarity 0.89)
- Let 2D Diffusion Model Know 3D-Consistency for Robust Text-to-3D Generation (ICLR 2024 · similarity 0.88)
- What Matters When Repurposing Diffusion Models for General Dense Perception Tasks? (ICLR 2025 · similarity 0.88)
- CONFORM: Contrast is All You Need for High-Fidelity Text-to-Image Diffusion Models (CVPR 2024 · similarity 0.87)
- Towards Understanding Cross and Self-Attention in Stable Diffusion for Text-Guided Image Editing (CVPR 2024 · similarity 0.87)
- $\text{I}^2\text{AM}$: Interpreting Image-to-Image Latent Diffusion Models via Bi-Attribution Maps (ICLR 2025 · similarity 0.87)
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.
