CrossHOI-Bench：跨视觉语言模型与 HOI 专用方法的统一 HOI 评测基准
CrossHOI-Bench: A Unified Benchmark for HOI Evaluation across Vision-Language Models and HOI-Specific Methods
用带显式正例与精选负例的多项选择协议替代 HICO-DET 的精确标签匹配，可直接用于公平比较 VLM 与专用 HOI 模型，并暴露多动作并发与动作-人物绑定这两类失败模式。
原文依据 we introduce CrossHOI-Bench, a multiple-choice HOI benchmark with explicit positives and curated negatives
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-物交互（HOI）检测长期由任务专用模型主导，早期有时还会使用 CLIP 等视觉语言骨干网络。随着大型生成式视觉语言模型（VLM）的兴起，一个关键问题是：独立的 VLM 能否在与专用 HOI 方法的竞争中完成 HOI 检测。HICO-DET 等现有基准要求在标注不完整的情况下进行精确标签匹配，因此任何未匹配的预测都会被判为错误。这会不公平地惩罚有效输出，尤其是来自约束较少的 VLM 的输出，并使跨范式比较不可靠。为解决这一局限，我们提出 CrossHOI-Bench，这是一个多项选择式 HOI 基准，包含显式正例与精选负例，可对 VLM 和 HOI 专用模型进行统一且可靠的评测。我们进一步聚焦多人场景和细粒度交互区分等具有挑战性的场景，这些场景对于揭示两种范式之间的真实差异至关重要。实验表明，大型 VLM 取得了有竞争力、有时甚至更优的零样本性能，但它们在处理多个并发动作以及将交互正确分配给目标人物方面仍存在困难。相反，HOI 专用方法在通用 HOI 推理上仍然较弱，但表现出更强的多动作识别能力，并能更可靠地识别哪个人执行了哪个动作。这些发现揭示了 VLM 与 HOI 专用方法互补的优势与劣势，而现有基准由于错误的惩罚机制无法揭示这些差异。
英文原文摘要 1523 字符
HOI detection has long been dominated by task-specific models, sometimes with early vision-language backbones such as CLIP. With the rise of large generative VLMs, a key question is whether standalone VLMs can perform HOI detection competitively against specialized HOI methods. Existing benchmarks such as HICO-DET require exact label matching under incomplete annotations, so any unmatched prediction is marked wrong. This unfairly penalizes valid outputs, especially from less constrained VLMs, and makes cross-paradigm comparison unreliable. To address this limitation, we introduce CrossHOI-Bench, a multiple-choice HOI benchmark with explicit positives and curated negatives, enabling unified and reliable evaluation of both VLMs and HOI-specific models. We further focus on challenging scenarios, such as multi-person scenes and fine-grained interaction distinctions, which are crucial for revealing real differences between the two paradigms. Experiments show that large VLMs achieve competitive, sometimes superior, zero-shot performance, yet they struggle with multiple concurrent actions and with correctly assigning interactions to the target person. Conversely, HOI-specific methods remain weaker in general HOI reasoning but demonstrate stronger multi-action recognition and more reliable identification of which person performs which action. These findings expose complementary strengths and weaknesses of VLMs and HOI-specific methods, which existing benchmarks fail to reveal due to incorrect penalization.