HimmBO: High-Dimensional Interactive Model-Merging Bayesian Optimization for Image Generation

概要

Fine-tuning has become a common way to customize text-to-image models, leading to community collections of over 100K lightweight adapters. Adapters sharing a base model can be merged with continuous coefficients. This turns a large collection into a high-dimensional continuous design space, which recent work explores interactively using preferential Bayesian optimization (PBO). However, existing approaches are limited to a few dozen preselected adapters and require expensive image generation in every interaction round. We introduce HimmBO, a two-stage method that scales preference-driven adapter merging to much larger collections while reducing image-generation cost. Its screening stage recovers the active adapter set through level-set estimation over pre-generated images, requiring no image generation. The refinement stage then optimizes merge coefficients in the recovered low-dimensional subspace. Both stages use a style-aware kernel, allowing preference information to transfer between visually similar adapters. In simulated-user image-matching experiments, HimmBO achieves higher final similarity than all comparisons while reaching a given similarity with under one tenth of the image generations required by prior work. Its performance remains stable as the collection grows to 1080 adapters, while competing methods degrade substantially.

参考リンク

• Paper ↗

リンク先のライセンス・再現性・データ条件をご確認ください。自動分類や実装の目安は保証ではありません。

実装を始めるプロンプト

参考リンクを確認したうえで、次の内容をコピーして利用できます。

# 論文・記事
HimmBO: High-Dimensional Interactive Model-Merging Bayesian Optimization for Image Generation

# 概要
Fine-tuning has become a common way to customize text-to-image models, leading to community collections of over 100K lightweight adapters. Adapters sharing a base model can be merged with continuous…

# 原典
https://arxiv.org/abs/2610.30875

上記の手法を実装してください。実行前に依存関係とライセンスを確認してください。