AI Paper Atlas

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

Mitigating Noise-Induced Layout Priors for Object Counting in Diffusion Models

Xiaoling Gu, Xuelong Li, Shengqi Wu, Yongkang Wong, Zizhao Wu, Huan Li, Zhou Yu, Mohan Kankanhalli

ICML 2026 · Image generation › Text-to-image diffusion · approach: Compositional text-to-image guidance

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 Object Saliency Enhancement (ablation)Ours (SDXL) · T2I-Countaccuracy: 0.5378 -> 0.5556Improved
add LRR and OSE (count-aware noise adjustment) (ablation)Ours (SDXL) · T2I-Countaccuracy: 0.5022 -> 0.5556Improved
add Attention Concentration Loss (ablation)Ours (SDXL) · T2I-Countaccuracy: 0.3622 -> 0.5556Improved
add Dispersion Regularization Loss (ablation)Ours (SDXL) · T2I-Countaccuracy: 0.5311 -> 0.5556Improved
add both attention losses (layout consistency strategy) (ablation)Ours (SDXL) · T2I-Countaccuracy: 0.3511 -> 0.5556Improved
add Attention Concentration Loss (ablation)Ours (SDXL) · T2I-CountSTA: 0.6701 -> 0.8123Improved
add Object Saliency Enhancement (ablation)Ours (SDXL) · T2I-CountMAE: 1.1267 -> 0.9889Improved
apply the proposed training-free counting frameworkSDXL · COCOCountaccuracy: 0.2958 -> 0.6125Improved
apply the proposed training-free counting frameworkSD v3 · COCOCountaccuracy: 0.5058 -> 0.6225Improved
apply the proposed training-free counting frameworkSDXL · T2I-Countaccuracy: 0.2978 -> 0.5556Improved
apply the proposed training-free counting frameworkSD v3 · T2I-Countaccuracy: 0.4444 -> 0.6022Improved
apply the proposed training-free counting frameworkSD v3 · MultiCountaccuracy: 0.1486 -> 0.3500Improved

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