Research map / Image generation / Autoregressive and latent generation
Latent-NeRF for Shape-Guided Generation of 3D Shapes and Textures
Read the paperCheck 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 |
|---|---|---|---|
| Render and optimize directly in latent space rather than encoding rendered RGB images at each guidance step | Latent-NeRF · Single V100 | convergence time: about 30 minutes -> less than 15 minutes | Helped |
| Add RGB-space refinement after latent-space training | Latent-NeRF · Complex objects or regions with detailed textures | visual detail quality | Helped |
| Add RGB-space refinement after latent-space training | Latent-NeRF · RGB refinement iterations | iteration speed: ×2 slower | Hurt |
| Add MeshSketch occupancy guidance to constrain the generated geometry | Latent-NeRF · MeshSketch-guided shape generation | control over generated shape | Helped |
| Set MeshSketch leniency parameter σ_S to 0.05 (ablation) | Latent-NeRF with MeshSketch guidance · MeshSketch-guided shape generation | shape resemblance to conditioning shape: σ_S=0.05 | Helped |
| Increase MeshSketch leniency parameter σ_S to 1.5 (ablation) | Latent-NeRF with MeshSketch guidance · MeshSketch-guided shape generation | shape resemblance to conditioning shape: σ_S=1.5 | Hurt |
| Use UV texture maps instead of optimizing per-face latent attributes (ablation) | Latent-Paint · Texture generation on 3D meshes | texture granularity independent of geometric resolution | Helped |
| Use diffusion-model guidance to generate mesh textures | Latent-Paint · Texture generation on 3D meshes | texture precision | Helped |
Closest papers by meaning
- HIFA: High-fidelity Text-to-3D Generation with Advanced Diffusion Guidance (ICLR 2024 · similarity 0.90)
- ED-NeRF: Efficient Text-Guided Editing of 3D Scene With Latent Space NeRF (ICLR 2024 · similarity 0.90)
- DreamTime: An Improved Optimization Strategy for Diffusion-Guided 3D Generation (ICLR 2024 · similarity 0.89)
- Consistent3D: Towards Consistent High-Fidelity Text-to-3D Generation with Deterministic Sampling Prior (CVPR 2024 · similarity 0.89)
- Text-to-3D with Classifier Score Distillation (ICLR 2024 · similarity 0.89)
- DreamCraft3D: Hierarchical 3D Generation with Bootstrapped Diffusion Prior (ICLR 2024 · similarity 0.89)
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.