探究数据驱动模型是否学到了雷达运动背后的物理规律。
What Physics do Data-Driven MoCap-to-Radar Models Learn?
- 用两个物理可解释性指标评估模型预测与真实多普勒频率的匹配度。
- 发现重建误差低的模型仍可能不符合物理规律，仅部分模型真正学习了物理。
- 揭示时间注意力机制对Transformer模型学习物理至关重要，适合研究模型可解释性者阅读。
数据驱动的动捕到雷达模型能生成合理的微多普勒谱图，但它们是否真正学到了底层物理规律？我们提出一种基于物理的可解释性框架，通过两个互补指标来回答该问题：第一个衡量模型预测与物理推导出的多普勒频率的一致性；第二个测试在速度干预下预测是否保持速度-频率关系。两个指标仅需动捕输入和模型输出，无需实际雷达数据。跨多种模型架构的实验表明，低重建误差并不保证物理一致性：部分模型虽误差低，但在两个物理指标上表现差。进一步分析显示，时间注意力对Transformer类模型学习底层物理至关重要。
原文摘要 · Abstract (English)
Data-driven MoCap-to-radar models generate plausible micro-Doppler spectrograms, but do they actually learn the underlying physics? We introduce a physics-based interpretability framework to answer this question via two proposed complementary metrics: one measures alignment between model predictions and the physics-derived Doppler frequency, while the other tests whether predictions preserve the velocity-frequency relationship under velocity intervention. Both metrics require only MoCap input and model predictions, without access to measured radar data. Experiments across several model architectures reveal that low reconstruction error does not guarantee physical consistency: some, but not all, models achieve low error yet perform poorly on the two physics-based metrics. Further analysis shows that temporal attention is critical for transformer-based models to learn the underlying physics.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。