AI Paper Atlas

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

Anchoring and Steering Diffusion: Enhancing the Faithfulness of Text-to-Image Generation at Inference Time

Xinyi Wang, Yuyang Huang, Yalin Su, Pengcheng Luan, Tao Zhang, FEIMING WEI, Wenxian Yu

ECCV 2026 · Image generation › Text-to-image diffusion · approach: Compositional text-to-image guidance · 0 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 Semantic Anchoring (ablation)SDXL · GenEvaloverall accuracy: 55.933 -> 60.731Improved
add Reflective Steering (ablation)SDXL · GenEvaloverall accuracy: 55.933 -> 59.868Improved
add Reflective Steering to Semantic Anchoring (ablation)SDXL · GenEvaloverall accuracy: 60.731 -> 63.393Improved
add Semantic Anchoring (ablation)HunyuanDiT-v1.2 · GenEvaloverall accuracy: 68.342 -> 70.395Improved
add Reflective Steering (ablation)HunyuanDiT-v1.2 · GenEvaloverall accuracy: 68.342 -> 71.728Improved
add Reflective Steering to Semantic Anchoring (ablation)HunyuanDiT-v1.2 · GenEvaloverall accuracy: 70.395 -> 72.470Improved
add LP-SDS refinement (N_lp=400) to CLIP-only prior (ablation)AnchorSteer on SDXL · GenEvaloverall accuracy: 60.768 -> 63.393Improved
increase LP-SDS steps from 400 to 600 (ablation)AnchorSteer on SDXL · GenEvaloverall accuracy: 63.393 -> 62.512Hurt
increase inspection stride from 1 to 5 (ablation)AnchorSteer on SDXL · GenEvaloverall accuracy: 63.393 -> 59.950Hurt
use a 2B VLM instead of Qwen-VL-Chat for diagnosis (ablation)AnchorSteer on SDXL · GenEvaloverall accuracy: 63.393 -> 62.073Mixed
apply AnchorSteer on top of standard samplingSDXL · GenEval image qualityImageReward: 0.4891 -> 0.8781Improved
add Semantic Anchoring (ablation)SDXL · GenEval sample diversity1-LPIPS (lower is more diverse): 0.3860 -> 0.4406Hurt

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