超越帧的感知：基于原始时序视频与多模态线索的零样本行人意图预测
Seeing Beyond Frames: Zero-Shot Pedestrian Intention Prediction with Raw Temporal Video and Multimodal Cues
AI总结：
针对传统行人意图预测方法需大量重训练适配新场景的问题，提出基于Gemini 2.5 Pro的零样本BF-PIP方法，通过处理连续时序视频并融合多模态上下文线索，无需额外训练便实现73%准确率，优于GPT-4V基线18%，为智能交通敏捷感知模块提供新思路。
摘要：
行人意图预测是复杂城市环境下自动驾驶的核心功能之一。传统方法依赖对帧序列的监督学习，需大量重新训练才能适配新场景。本文提出BF-PIP（超越帧行人意图预测），这是一种基于Gemini 2.5 Pro构建的零样本方法，可直接从补充了结构化JAAD元数据的连续短视频片段推断行人过街意图。与基于GPT-4V、处理离散帧的方法不同，BF-PIP处理无中断的时序片段，还通过专门的多模态提示融入了边界框标注和自车速度信息。在无需额外训练的情况下，BF-PIP达到73%的预测准确率，比GPT-4V基线高出18%。这些结果表明，将时序视频输入与上下文线索结合可增强时空感知能力，提升模糊场景下的意图推断效果。该方法为智能交通系统中无需重新训练的敏捷感知模块奠定了基础。
英文摘要：
Pedestrian intention prediction is essential for autonomous driving in complex urban environments. Conventional approaches depend on supervised learning over frame sequences and require extensive retraining to adapt to new scenarios. Here, we introduce BF-PIP (Beyond Frames Pedestrian Intention Prediction), a zero-shot approach built upon Gemini 2.5 Pro. It infers crossing intentions directly from short, continuous video clips enriched with structured JAAD metadata. In contrast to GPT-4V based methods that operate on discrete frames, BF-PIP processes uninterrupted temporal clips. It also incorporates bounding-box annotations and ego-vehicle speed via specialized multimodal prompts. Without any additional training, BF-PIP achieves 73% prediction accuracy, outperforming a GPT-4V baseline by 18 %. These findings illustrate that combining temporal video inputs with contextual cues enhances spatiotemporal perception and improves intent inference under ambiguous conditions. This approach paves the way for agile, retraining-free perception module in intelligent transportation system.
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。
