AI Paper Atlas

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

InnoAds-Composer: Efficient Condition Composition for E-Commerce Poster Generation

Yuxin Qin, Ke Cao, Haowei Liu, Ao Ma, Fengheng Li, Honghe Zhu, Zheng Zhang, Run Ling, Wei Feng, Xuanhua He, Zhanjie Zhang, Zhen Guo and 4 more

CVPR 2026 · Image generation › Text-to-image diffusion · approach: Layout-to-image control · 1 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
add Stage II token pruning and fine-tuningInnoAds-Composer · Inference efficiencyLatency (s): 55.87 -> 47.32Improved
add Stage II token pruning and fine-tuningInnoAds-Composer · InnoComposer-BenchFID: 54.39 -> 55.24Hurt
add Stage II token pruning and fine-tuningInnoAds-Composer · InnoComposer-BenchSen. Acc: 0.857 -> 0.847Hurt
add Stage II token pruning and fine-tuningInnoAds-Composer · Inference efficiencyFLOPs (T): 165.56 -> 135.25Improved
add decoupled attention (vs native full attention)InnoAds-Composer Stage I vs Flux-Kontext · Inference efficiencyLatency (s): 76.02 -> 55.87Improved
add Text Feature Enhancement Module (ablation)InnoAds-Composer · InnoComposer-BenchSen. Acc: approximately 5% improvementImproved
use importance-aware instead of random or uniform token pruning (ablation)InnoAds-Composer Stage I · InnoComposer-BenchNED, MSE, CLIP-IImproved
prune glyph/subject/style tokens at 90/60/70 instead of 80/50/60 (ablation)InnoAds-Composer Stage II · Stage II token pruning ratiosNED: 0.969 -> 0.582Hurt
prune glyph/subject/style tokens at 50/20/30 instead of 80/50/60 (ablation)InnoAds-Composer Stage II · Stage II token pruning ratiosNED: 0.971 vs 0.969No clear effect
prune at 90/60/70 instead of 80/50/60 (ablation)InnoAds-Composer Stage II · Stage II token pruning ratiosCLIP-I: 0.594 -> 0.451Hurt

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most 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

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