Research map / 3D vision / Face and shape
DreamStone: Image as a Stepping Stone for Text-Guided 3D Shape Generation
Zhengzhe Liu, Peng Dai, Ruihui Li, Xiaojuan Qi, Chi‐Wing Fu
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 stage-1 image-to-shape alignment (ablation) | DreamStone · ShapeNet text-to-shape | FID: 222.96 -> 114.34 | Improved |
| add stage-1 image-to-shape alignment (ablation) | DreamStone · ShapeNet text-to-shape | Consistency Score (%): 1.92 -> 70.77 | Improved |
| add stage-2 text-to-shape alignment (ablation) | DreamStone · ShapeNet text-to-shape | FID: 202.33 -> 114.34 | Improved |
| add stage-2 text-to-shape alignment (ablation) | DreamStone · ShapeNet text-to-shape | Consistency Score (%): 29.52 -> 70.77 | Improved |
| add background loss in stage 1 (ablation) | DreamStone · ShapeNet text-to-shape | FID: 149.45 -> 114.34 | Improved |
| add background loss in stage 2 (ablation) | DreamStone · ShapeNet text-to-shape | FID: 156.52 -> 114.34 | Improved |
| add background loss in both stages (ablation) | DreamStone · ShapeNet text-to-shape | FID: 178.34 -> 114.34 | Improved |
| add background loss in both stages (ablation) | DreamStone · ShapeNet text-to-shape | Consistency Score (%): 30.96 -> 70.77 | Improved |
| map CLIP image feature to SVR shape space instead of decoding from CLIP image encoder (ablation) | DVR decoder · ShapeNet text-to-shape | FID: 181.88 -> 114.34 | Improved |
| use mapped two-stage alignment instead of text-to-image then DVR (LAFITE+DVR) | DreamStone · ShapeNet text-to-shape | FID: 135.01 -> 114.34 | Improved |
| add diffusion prior and SDS refinement to ISS | ISS · ShapeNet text-to-shape | FID: 124.42 -> 114.34 | Improved |
| add diffusion prior and SDS refinement to ISS | ISS · ShapeNet text-to-shape | Consistency Score (%): 60.0 -> 70.77 | Improved |
Closest papers by meaning
- ISS: Image as Stepping Stone for Text-Guided 3D Shape Generation (ICLR 2023 · similarity 0.94)
- Dream3D: Zero-Shot Text-to-3D Synthesis Using 3D Shape Prior and Text-to-Image Diffusion Models (CVPR 2023 · similarity 0.90)
- ShapeWords: Guiding Text-to-Image Synthesis with 3D Shape-Aware Prompts (CVPR 2025 · similarity 0.86)
- PI3D: Efficient Text-to-3D Generation with Pseudo-Image Diffusion (CVPR 2024 · similarity 0.86)
- ViewDiff: 3D-Consistent Image Generation with Text-to-Image Models (CVPR 2024 · similarity 0.85)
- Vector Grimoire: Codebook-based Shape Generation under Raster Image Supervision (ICML 2025 · similarity 0.84)
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.
