Research map / Image generation / Text-to-image diffusion
TaxaDiffusion: Progressively Trained Diffusion Model for Fine-Grained Species Generation
Amin Karimi Monsefi, Mridul Khurana, Rajiv Ramnath, Anuj Karpatne, Wei-Lun Chao, Cheng Zhang
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 progressive taxonomy training instead of all levels at once (ablation) | TaxaDiffusion · FishNet | BioCLIP score: 9.32 -> 10.43 | Improved |
| use progressive taxonomy training instead of all levels at once (ablation) | TaxaDiffusion · FishNet | LPIPS: 0.7487 -> 0.7319 | Improved |
| use progressive taxonomy training instead of all levels at once (ablation) | TaxaDiffusion · FishNet | FID: 32.43 -> 31.87 | Improved |
| use progressive training instead of random level selection (ablation) | TaxaDiffusion · FishNet | FID: 29.53 -> 31.87 | Hurt |
| use progressive training instead of random level selection (ablation) | TaxaDiffusion · FishNet | BioCLIP score: 9.85 -> 10.43 | Improved |
| replace unconditional guidance with Kingdom-level conditioning (TaxaGuide) (ablation) | TaxaDiffusion · FishNet | FID: 47.61 -> 32.42 | Improved |
| replace unconditional guidance with Kingdom-level conditioning (TaxaGuide) (ablation) | TaxaDiffusion · FishNet | BioCLIP score: 9.45 -> 11.53 | Improved |
| replace unconditional guidance with Kingdom-level conditioning (TaxaGuide) (ablation) | TaxaDiffusion · FishNet | LPIPS: 0.7313 -> 0.7092 | Improved |
| add progressive taxonomy conditioning on top of LoRA | SD + LoRA · FishNet Species level | FID: 43.91 -> 31.87 | Improved |
| add progressive taxonomy conditioning versus full fine-tuning | SD + Full · FishNet Species level | BioCLIP score: 8.31 -> 10.43 | Improved |
| use TaxaDiffusion instead of FineDiffusion (DiT-XL/2) | FineDiffusion · iNaturalist 171 shared species | FID: 74.80 -> 43.71 | Improved |
| use TaxaDiffusion instead of zero-shot SD | SD · iNaturalist Species level | FID: 73.13 -> 46.39 | Improved |
Closest papers by meaning
- Learning the 3D Fauna of the Web (CVPR 2024 · similarity 0.87)
- BioCLIP 2: Emergent Properties from Scaling Hierarchical Contrastive Learning (NeurIPS 2025 · similarity 0.86)
- BioTrove: A Large Curated Image Dataset Enabling AI for Biodiversity (NeurIPS 2024 · similarity 0.85)
- What Do You See in Common? Learning Hierarchical Prototypes over Tree-of-Life to Discover Evolutionary Traits (ICLR 2025 · similarity 0.85)
- Reconstructing Animals and the Wild (CVPR 2025 · similarity 0.84)
- Dis²Booth: Learning Image Distribution with Disentangled Features for Text-to-Image Diffusion Models (AAAI 2025 · 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.
