Research map / 3D vision / 3D generation
Tree-D Fusion: Simulation-Ready Tree Dataset from Single Images with Diffusion Priors
Jae Joong Lee, Bosheng Li, Sara M Beery, Jonathan Huang, Songlin Fei, Raymond A. Yeh, Bedrich Benes
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 |
|---|---|---|---|
| use tree-specific diffusion priors (LoRA 2D plus tree-trained Zero123 3D) with genus-conditioned space colonization instead of generic single-image 3D methods | Tree-D Fusion · AAD San Jose | ICTree realism: 0.45 -> 0.71 | Improved |
| use Tree-D Fusion instead of RBV | Tree-D Fusion · AAD San Jose | LPIPS: 0.58 -> 0.54 | Improved |
| use Tree-D Fusion instead of RBV | Tree-D Fusion · AAD San Jose | CLIP similarity: 0.54 -> 0.63 | Improved |
| use Tree-D Fusion instead of best baseline | Tree-D Fusion · LiDAR trees | Chamfer distance (1e-2): 4.25 -> 3.25 | Improved |
| Tree-D Fusion versus RBV on ICTree | Tree-D Fusion · Cupressus (ICTree) | ICTree realism: 0.75 -> 0.65 | Hurt |
| set prior weight ratio alpha/beta to 0.1 instead of 0.01 (ablation) | Tree-D Fusion · AAD San Jose | CLIP similarity: 0.53 -> 0.60 | Improved |
| raise alpha/beta from 0.1 to 10 (more weight on 2D prior) (ablation) | Tree-D Fusion · AAD San Jose | ICTree realism: 0.55 -> 0.47 | Hurt |
| replace NeRF-style diffusion-prior reconstruction with 3D Gaussian Splatting novel view synthesis (ablation) | Tree reconstruction · tree novel view synthesis | qualitative | Hurt |
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.
- Magic123 — Magic123: One Image to High-Quality 3D Object Generation Using Both 2D and 3D Diffusion Priors (ICLR 2024) · better on 4 of 4 shared results
- DreamGaussian — DreamGaussian: Generative Gaussian Splatting for Efficient 3D Content Creation (ICLR 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
- SVDTree: Semantic Voxel Diffusion for Single Image Tree Reconstruction (CVPR 2024 · similarity 0.91)
- OAM-TCD: A globally diverse dataset of high-resolution tree cover maps (NeurIPS 2024 · similarity 0.85)
- Diffusion Tree Sampling: Scalable inference‑time alignment of diffusion models (NeurIPS 2025 · similarity 0.85)
- LT3SD: Latent Trees for 3D Scene Diffusion (CVPR 2025 · similarity 0.84)
- Neural Hierarchical Decomposition for Single Image Plant Modeling (CVPR 2025 · similarity 0.83)
- Geometry-Centered 3D Latent World Models for Growing Surfaces (NeurIPS 2026 · similarity 0.83)
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.
