AI画画终于能分清“风格”和“内容”了
现在的AI画画，你给一张“风格图”和一张“内容图”，它常常把风格图里的物体也带进来，比如风格图里有朵花，生成图里就莫名其妙多出朵花。这篇论文把这个问题拆成两步：先让模型学会只迁移风格（比如笔触、色调），再学会同时保留内容的结构和语义。他们从社区LoRA模型里挖出大量干净的数据对，还设计了一个“注意力约束”和“频率调制”来堵住风格泄露的漏洞。最终模型在风格相似度、内容保留和防泄露上都比现有方法强。它不是你明天就能用的工具，但让“风格迁移”这个老问题有了更可靠的解法。
📄 原文摘要(英文)
Style-content dual-reference generation aims to synthesize an image that preserves the structure and semantics of a content reference while adopting the style of a separate style reference.Despite recent progress, this setting remains challenging because models must balance content fidelity, style alignment, and instruction following avoiding semantic leakage from the style reference.A key bottleneck is the lack of large-scale triplet data with clean content-style separation and broad long-tail style coverage.In this work, we propose FreeStyle, a scalable dual-reference generation framework based on community LoRA mining.We treat community LoRAs as compositional anchors for style and content, and design a rigorous generation and filtering pipeline to construct large-scale Style-Reference and Content-Reference triplets across multiple base models.To address content leakage, we adopt a two-stage curriculum with stage-specific disentanglement mechanisms: an attention-level enrichment constraint that suppresses style-reference leakage in the style-transfer stage, and a frequency-aware RoPE modulation strategy that targets positional-correspondence-based leakage in the harder dual-reference stage.We also introduce a benchmark covering both style-reference and dual-reference generation, with evaluations on style similarity, content preservation, aesthetics, instruction following, and leakage rejection. The benchmark incorporates a style-invariant Content Alignment Score (CAS) and introduces a calibrated VLM-based Rejection Score for evaluating generation reliability and leakage suppression.Extensive experiments show that our model achieves a strong balance among style alignment, content preservation, and leakage suppression.