AI Paper Atlas

Research map / Image generation / Text-to-image diffusion

PEA-Diffusion: Parameter-Efficient Adapter with Knowledge Distillation in non-English Text-to-Image Generation

jian ma, Chen Chen, Qingsong Xie, Haonan Lu

ECCV 2024 · Image generation › Text-to-image diffusion · approach: Compositional text-to-image guidance · 12 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
train a 6M-parameter MLP adapter with UNet distillation instead of translationPEA-Diffusion (SDXL) · MC cultural prompts, user studysuccess rate: 0.2577 -> 0.4978Improved
use PEA-Diffusion instead of translation on general promptsPEA-Diffusion (SDXL) · MG general promptsImageReward / PickScoreHurt
use PEA-Diffusion instead of translation for JapanesePEA-Diffusion · MC cultural prompts, JapaneseCLIPScoreHurt
add KD of any kind versus plain fine-tuning of the adapter (ablation)PEA-Diffusion · Chinese MG/MCImageReward, PickScore, CLIPScoreImproved
add UNet intermediate feature distillation (ablation)PEA-Diffusion · Chinese MG/MCImageReward, CLIPScoreImproved
add UNet logit distillation only (ablation)PEA-Diffusion · Chinese MG/MCImageReward, CLIPScoreNo clear effect
distill text encoder features only (TKD) (ablation)PEA-Diffusion · Chinese MG/MCImageReward, CLIPScoreHurt
add text encoder KD on top of UNet KD (ablation)PEA-Diffusion · Chinese MG/MCImageReward, CLIPScoreHurt
replace adapter training with LoRA or W_K/W_V fine-tuningSDXL language transfer · MG and MCImageReward, PickScore, CLIPScoreHurt
use PEA-Diffusion instead of GlueGenPEA-Diffusion · MG, five languagesImageReward: GlueGen -0.5 ImageRewardImproved
train PEA on low-resolution (around 640) data (ablation)PEA-Diffusion (SDXL, Chinese) · MGImageRewardHurt
increase adapter parameters at equal training samples (ablation)PEA · MG (Chinese)ImageRewardImproved

Later papers that compared with this method

1 later accepted paper reports PEA-Diffusion as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.

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

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