LensErase:可擦除任意内容的基础模型
原标题:Paper page - LensErase: A Foundation Model for Erase Anything
AI 摘要
研究团队提出 LensErase,一个基于掩码的通用「擦除任意内容」基础模型,覆盖物体、视觉文字、水印、视觉特效及多目标移除等多种真实场景。为支撑广泛覆盖,团队构建了包含 1.0107 亿对源图-掩码-目标样本的大规模数据集 LensErase-100M,并从文本到图像模型 Lens 预训练,提出 Mask-Aware RoPE 将掩码信息直接注入源图 token。模型经监督微调与带合约引导、区域感知奖励的强化学习精调,推理时使用 TraceFusion 保留编辑区域外的高保真内容,在六个公开基准及自建 LensErase-Bench 上取得显著提升。
正文节选
LensErase: A Foundation Model for Erase Anything Abstract We introduce LensErase, a general-purpose mask-based Erase Anything foundation model designed to cover a broad spectrum of real-world erase scenarios. LensErase supports diverse objects, visual text, watermarks, associated visual effects, multiple-object removal, and a wide range of object scales, aspect ratios, and image resolutions. First, to enable such broad coverage, we construct LensErase-100M, a large-scale dataset containing 101.0