@inproceedings{6a7a92f462614f2b8054073cdd4abcc9,
title = "An auxiliary training method for single-player badminton",
abstract = "Aiming at the simulation training of badminton movement, this paper proposes a method of adjusting the deflection angle of the limbs based on monocular video movement reconstruction. The confidence and recognition rate were used to judge the effect of gesture recognition by detecting the key joints and their connections. We extracted depth information based on a residual learning full convolutional network and realized monocular video depth estimation. We also designed a neural network regressor to reconstruct 3D motions. The method was verified by comparing the motion capture data with 3D motion data generated by the monocular video.",
keywords = "Auxiliary training, Deep learning, Monocular 3D motion reconstruction, Motion capture",
author = "Tianyu Huang and Yihao Li and Wentao Zhu",
note = "Publisher Copyright: {\textcopyright} 2021 IEEE.; 16th IEEE International Conference on Computer Science and Education, ICCSE 2021 ; Conference date: 17-08-2021 Through 21-08-2021",
year = "2021",
month = aug,
day = "17",
doi = "10.1109/ICCSE51940.2021.9569592",
language = "English",
series = "ICCSE 2021 - IEEE 16th International Conference on Computer Science and Education",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
pages = "441--446",
booktitle = "ICCSE 2021 - IEEE 16th International Conference on Computer Science and Education",
address = "United States",
}