自主监督学习提高了基于IMU的地面反应力估计数据的准确性和效率
Tian Tan1, Peter B Shull2, Jenifer L Hicks3
1Department of Radiology, Stanford University, Stanford, CA, 94305, USA.
bioRxiv : the preprint server for biology
|February 8, 2024
概括
自主监督学习 (SSL) 使用惯性测量单位 (IMU) 数据显著提高地面反应力 (GRF) 估计准确度. 这种方法提高了数据效率,减少了在动力评估中需要广泛的标记GRF数据的需求.
科学领域:
- 生物力学 生物力学
- 机器学习 机器学习
- 可穿戴技术可穿戴技术
背景情况:
- 用于惯性测量单元 (IMU) 驱动的动力评估的深度学习需要大量的地面反应力 (GRF) 数据来进行监督训练.
- 这种依赖标记的GRF数据对实际应用构成了瓶.
研究的目的:
- 调查自主监督学习 (SSL) 技术的有效性,用于使用大型IMU数据集预训练深度学习模型.
- 提高基于IMU的GRF估计的准确性和数据效率.
主要方法:
- 通过掩盖IMU数据和训练变压器模型重建掩盖部分来执行SSL.
- 该研究比较了真实,合成和组合IMU数据集中的各种掩盖比率.
- 模型在地面行走,跑步机行走和落地任务中对GRF估计进行了评估.
主要成果:
- 与使用相同数量的标记数据的监督方法相比,SSL预训练显著提高了步行期间3轴GRF估计准确度.
- 用仅1-10%的行走数据微调SSL模型,实现了与100%数据训练的基线模型可比的准确性.
- 对SSL的最佳掩护比率被确定为6.25-12.5%.
结论:
- SSL有效地利用大型IMU数据集 (真实和合成) 来促进基于深度学习的GRF估计.
- 这种方法大大减少了对标记的GRF数据的要求,使IMU驱动的动力评估更容易获得.
相关概念视频
Application of Linearization and Approximation
A drone flying through complex terrain often relies on more than one sensing method to estimate small changes in altitude. Along with direct measurements, air pressure provides a useful indirect indicator of vertical movement. Atmospheric pressure decreases as altitude increases, and this relationship is commonly described using an exponential model. Although accurate, converting pressure measurements into altitude values requires calculations that are too complex to perform repeatedly during...
Vector Functions and Motion: Problem Solving
Accurate position tracking is fundamental to the safe and effective operation of unmanned aerial vehicles (UAVs), particularly during precision maneuvers near complex structures. In this scenario, a drone is programmed to perform a high-precision inspection of a vertical structure, starting at position ((x, y, z) = (3, 0, 0)), with an initial velocity oriented in the positive z-direction. The trajectory of the drone is governed by a time-dependent acceleration function a(t), which is predefined...


