Research map / 3D vision / 3D generation
Zero-Shot Text-Guided Object Generation With Dream Fields
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 transmittance loss together with background augmentation (ablation) | Dream Fields · Object Centric COCO captions | R-Precision: +18% and +15.6% | Helped |
| Use the residual, normalized bottleneck MLP architecture (ablation) | Dream Fields · Object Centric COCO captions | R-Precision: +8% and +2% | Helped |
| Bound the scene to a cube (ablation) | Dream Fields · Object Centric COCO captions | R-Precision: +13% and +11% | Helped |
| Add transmittance loss without background augmentation (ablation) | Dream Fields · Object Centric COCO captions | R-Precision (CLIP B/32): 31.3725 -> 23.6842 | Hurt |
| Add random crop augmentation (ablation) | Dream Fields · Object Centric COCO captions | R-Precision (CLIP B/32): 31.3725 -> 25.8170 | Hurt |
| Add background augmentation (ablation) | Dream Fields · Object Centric COCO captions | R-Precision (CLIP B/32): 31.3725 -> 44.1176 | Helped |
| Use mip-NeRF integrated positional encoding (ablation) | Dream Fields · Object Centric COCO captions | R-Precision: 31.3725 -> 29.7386 | No clear effect |
| Use higher-frequency Fourier features (ablation) | Dream Fields · Object Centric COCO captions | R-Precision (CLIP B/32): 31.3725 -> 24.1830 | Hurt |
| Use clipped additive transmittance regularization instead of no regularizer (ablation) | Dream Fields · Object Centric COCO captions | R-Precision: 35.2941 -> 62.0915 | Helped |
| Use density perturbation (ablation) | Dream Fields · Object Centric COCO captions | R-Precision: +12.4% | Helped |
| Use a beta prior on transmittance (ablation) | Dream Fields · Object Centric COCO captions | R-Precision: +15% | Helped |
| Use un-clipped multiplicative transmittance gating (ablation) | Dream Fields · Object Centric COCO captions | R-Precision: 35.2941 -> 34.6405 | Hurt |
Closest papers by meaning
- DIRECT-3D: Learning Direct Text-to-3D Generation on Massive Noisy 3D Data (CVPR 2024 · similarity 0.87)
- Let 2D Diffusion Model Know 3D-Consistency for Robust Text-to-3D Generation (ICLR 2024 · similarity 0.87)
- ViewDiff: 3D-Consistent Image Generation with Text-to-Image Models (CVPR 2024 · similarity 0.87)
- TextNeRF: A Novel Scene-Text Image Synthesis Method based on Neural Radiance Fields (CVPR 2024 · similarity 0.86)
- LLM Blueprint: Enabling Text-to-Image Generation with Complex and Detailed Prompts (ICLR 2024 · similarity 0.86)
- DreamText: High Fidelity Scene Text Synthesis (CVPR 2025 · similarity 0.86)
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.