AI Paper Atlas

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

Personalized Image Generation via Human-in-the-loop Bayesian Optimization

Rajalaxmi Rajagopalan, Debottam Dutta, Yu-Lin Wei, Romit Roy Choudhury

ICML 2026 · Image generation › Text-to-image diffusion · approach: T2I evaluation and controllable adapters · 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
use multi-choice (multiwise) preference likelihood instead of pairwise (ablation)MultiBO · target alignmentCLIP-I: 0.81 -> 0.94Improved
use multi-choice (multiwise) preference likelihood instead of pairwise (ablation)MultiBO · target alignmentLPIPS (lower better): 0.68 -> 0.55Improved
increase choice set size from K=4 to K=6 (ablation)MultiBO · target alignmentLPIPS (lower better): 0.55 -> 0.54No clear effect
increase choice set size from K=4 to K=10 (ablation)MultiBO · target alignmentLPIPS (lower better): 0.55 -> 0.64Hurt
optimize in attention space instead of 100D latent subspace of x_t (ablation)MultiBO · target alignmentCLIP-I: 0.8014 -> 0.9364Improved
optimize in attention space instead of 100D latent subspace of x_t (ablation)MultiBO · target alignmentLPIPS (lower better): 0.6726 -> 0.5497Improved
use human scorer instead of CLIP-I2I metric scorer (ablation)MultiBO · target alignmentCLIP-I: 0.9246 -> 0.9364Improved
use human scorer instead of LPIPS metric scorer (ablation)MultiBO · target alignmentLPIPS (lower better): 0.5924 -> 0.5497Improved
edit middle to later self-attention layers (34-64) rather than layers 1-24 (ablation)MultiBO · SDXL attention layer choicequalitativeImproved
edit attention only in first 20% of steps rather than whole trajectory (ablation)MultiBO · timestep rangequalitativeImproved
use multi-choice human BO instead of DEMON choose generate modeDEMON_choose · human evaluationwin rate (%): 15.60 -> 70.82Improved
use MultiBO with human scorer instead of Aesthetic reward scorerMultiBO_Aesthetic · human evaluationwin rate (%): 11.26 -> 70.82Improved

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