AI Paper Atlas

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

ICCV 2025 · Image generation › Text-to-image diffusion · approach: Text-driven synthesis and semantic guidance · 3 citations (OpenAlex)

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 changedWhereMeasuredEffect
use progressive taxonomy training instead of all levels at once (ablation)TaxaDiffusion · FishNetBioCLIP score: 9.32 -> 10.43Improved
use progressive taxonomy training instead of all levels at once (ablation)TaxaDiffusion · FishNetLPIPS: 0.7487 -> 0.7319Improved
use progressive taxonomy training instead of all levels at once (ablation)TaxaDiffusion · FishNetFID: 32.43 -> 31.87Improved
use progressive training instead of random level selection (ablation)TaxaDiffusion · FishNetFID: 29.53 -> 31.87Hurt
use progressive training instead of random level selection (ablation)TaxaDiffusion · FishNetBioCLIP score: 9.85 -> 10.43Improved
replace unconditional guidance with Kingdom-level conditioning (TaxaGuide) (ablation)TaxaDiffusion · FishNetFID: 47.61 -> 32.42Improved
replace unconditional guidance with Kingdom-level conditioning (TaxaGuide) (ablation)TaxaDiffusion · FishNetBioCLIP score: 9.45 -> 11.53Improved
replace unconditional guidance with Kingdom-level conditioning (TaxaGuide) (ablation)TaxaDiffusion · FishNetLPIPS: 0.7313 -> 0.7092Improved
add progressive taxonomy conditioning on top of LoRASD + LoRA · FishNet Species levelFID: 43.91 -> 31.87Improved
add progressive taxonomy conditioning versus full fine-tuningSD + Full · FishNet Species levelBioCLIP score: 8.31 -> 10.43Improved
use TaxaDiffusion instead of FineDiffusion (DiT-XL/2)FineDiffusion · iNaturalist 171 shared speciesFID: 74.80 -> 43.71Improved
use TaxaDiffusion instead of zero-shot SDSD · iNaturalist Species levelFID: 73.13 -> 46.39Improved

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy