OneHOI：统一人体-物体交互生成与编辑
OneHOI: Unifying Human-Object Interaction Generation and Editing
把人体-物体交互建模成角色/实例感知的 HOI token 加布局空间锚定，并用结构化注意力与 HOI RoPE 约束交互拓扑，这套条件注入方式可直接迁移到人物-物体合成、场景布局可控生成等需要多实体关系约束的扩散采样任务。
原文依据 a unified diffusion transformer framework that consolidates HOI generation and editing into a single conditional denoising process
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论人体-物体交互（HOI）建模刻画人如何作用于物体并与之产生关联，通常表示为<人, 动作, 物体>三元组。现有方法分为两个互不相通的分支：HOI 生成从结构化三元组与布局合成场景，但无法融合 HOI 与仅含物体实体等混合条件；HOI 编辑通过文本修改交互，却难以将姿态与物理接触解耦，也难以扩展到多个交互。我们提出 OneHOI，一个统一的扩散 Transformer 框架，将 HOI 生成与编辑整合为由共享结构化交互表示驱动的单一条件去噪过程。其核心是关系扩散 Transformer（R-DiT），它通过角色与实例感知的 HOI token 建模以动词为中介的关系，采用基于布局的空间动作锚定、强制交互拓扑的结构化 HOI 注意力，以及解耦多 HOI 场景的 HOI RoPE。结合模态丢弃在我们的 HOI-Edit-44K 以及 HOI 与以物体为中心的数据集上联合训练，OneHOI 支持布局引导、无布局、任意掩码与混合条件控制，在 HOI 生成与编辑两方面均取得最优结果。
英文原文摘要 1242 字符
Human-Object Interaction (HOI) modelling captures how humans act upon and relate to objects, typically expressed as <person, action, object> triplets. Existing approaches split into two disjoint families: HOI generation synthesises scenes from structured triplets and layout, but fails to integrate mixed conditions like HOI and object-only entities; and HOI editing modifies interactions via text, yet struggles to decouple pose from physical contact and scale to multiple interactions. We introduce OneHOI, a unified diffusion transformer framework that consolidates HOI generation and editing into a single conditional denoising process driven by shared structured interaction representations. At its core, the Relational Diffusion Transformer (R-DiT) models verb-mediated relations through role- and instance-aware HOI tokens, layout-based spatial Action Grounding, a Structured HOI Attention to enforce interaction topology, and HOI RoPE to disentangle multi-HOI scenes. Trained jointly with modality dropout on our HOI-Edit-44K, along with HOI and object-centric datasets, OneHOI supports layout-guided, layout-free, arbitrary-mask, and mixed-condition control, achieving state-of-the-art results across both HOI generation and editing.