Korol Anna S, Rodzin Taras, Zabava Kateryna, Gritsenko Valeriya
Annu Int Conf IEEE Eng Med Biol Soc. 2024 Jul;2024:1-6. doi: 10.1109/EMBC53108.2024.10782521.
Motion capture is used for recording complex human movements that is increasingly applied in medicine. We describe a novel algorithm of combining a machine learning approach with biomechanics to enable fast and robust analysis of motion capture data to obtain joint angles.
A multilayer perceptron and a recurrent neural network were compared in their capacity to estimate the joint angles of the human arm. The networks were pre-trained using data from a kinematic model of the human arm. The data comprised movements of three degrees of freedom, such as wrist flexion/extension, wrist ulnar/radial deviation, and hand pronation/supination.
A recurrent neural network model with long short-term memory architecture can solve the inverse kinematics problem for three rotational degrees of freedom with the least error; it performed faster than real time. The predictions were robust against noise.
This shows that it is feasible to rely on pre-trained neural networks for real-time calculation of joint angles.
运动捕捉用于记录复杂的人体运动，其在医学中的应用日益广泛。我们描述了一种将机器学习方法与生物力学相结合的新算法，以实现对运动捕捉数据的快速且稳健的分析，从而获得关节角度。
比较了多层感知器和循环神经网络估计人体手臂关节角度的能力。这些网络使用来自人体手臂运动学模型的数据进行预训练。数据包括三个自由度的运动，如手腕屈伸、手腕尺偏/桡偏以及手部旋前/旋后。
具有长短期记忆架构的循环神经网络模型能够以最小误差解决三个旋转自由度的逆运动学问题；其运行速度比实时速度还快。预测对噪声具有鲁棒性。
这表明依靠预训练的神经网络进行关节角度的实时计算是可行的。
