ScriptHOI：学习脚本化状态转换用于开放词汇人-物交互检测
ScriptHOI: Learning Scripted State Transitions for Open-Vocabulary Human-Object Interaction Detection
AI总结：
提出ScriptHOI框架，将交互短语分解为软脚本化状态转换，通过视觉状态分词器和槽位匹配器校准HOI逻辑，并引入区间部分标签学习和反事实脚本对比损失，提升开放词汇HOI检测中稀有和未见交互的识别，减少功能冲突误报。
摘要：
开放词汇人-物交互（HOI）检测需要识别在训练期间可能未作为注释类别出现的交互短语。最近的视觉-语言HOI检测器通过将人-物特征与文本嵌入匹配来改进语义迁移，但其预测通常受物体功能性和短语级共现主导。因此，模型可能仅凭刀和蛋糕的存在就预测“切蛋糕”，而未验证手、工具、目标、接触模式和物体状态是否共同支持该动作。我们提出 extbf{ScriptHOI}，一个结构化框架，将每个交互短语表示为软脚本化状态转换。ScriptHOI不将短语视为单个类别标记，而是将其分解为身体角色、接触、几何、功能性、运动和物体状态槽位。视觉状态分词器将每个检测到的人-物对解析为相应的状态标记，槽位匹配器估计脚本覆盖率和脚本冲突。这两个量校准HOI逻辑值，暴露缺失的视觉证据，并为不完整注释提供训练约束。为避免抑制有效但未注释的交互，我们进一步引入区间部分标签学习，该学习使用脚本导出的下界和上界概率约束未注释的候选，而不是分配封闭世界的负例。反事实脚本对比损失交换单个脚本槽位以阻止仅物体捷径。在HICO-DET、V-COCO和开放词汇HOI分割上的实验表明，ScriptHOI改善了稀有和未见交互的识别，同时大幅减少了功能冲突假阳性。
英文摘要：
Open-vocabulary human-object interaction (HOI) detection requires recognizing interaction phrases that may not appear as annotated categories during training. Recent vision-language HOI detectors improve semantic transfer by matching human-object features with text embeddings, but their predictions are often dominated by object affordance and phrase-level co-occurrence. As a result, a model may predict \textit{cut cake} from the presence of a knife and a cake without verifying whether the hand, tool, target, contact pattern, and object state jointly support the action. We propose \textbf{ScriptHOI}, a structured framework that represents each interaction phrase as a soft scripted state transition. Rather than treating a phrase as a single class token, ScriptHOI decomposes it into body-role, contact, geometry, affordance, motion, and object-state slots. A visual state tokenizer parses each detected human-object pair into corresponding state tokens, and a slot-wise matcher estimates both script coverage and script conflict. These two quantities calibrate HOI logits, expose missing visual evidence, and provide training constraints for incomplete annotations. To avoid suppressing valid but unannotated interactions, we further introduce interval partial-label learning, which constrains unannotated candidates with script-derived lower and upper probability bounds instead of assigning closed-world negatives. A counterfactual script contrast loss swaps individual script slots to discourage object-only shortcuts. Experiments on HICO-DET, V-COCO, and open-vocabulary HOI splits show that ScriptHOI improves rare and unseen interaction recognition while substantially reducing affordance-conflict false positives.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
