Selective Use of Skeletal Information to Reduce Computational Complexity of Motion Matching
Shohei Adachi, Ryohei Osawa, Hiroshi K. WATANABE · 2022 IEEE 11th Global Conference on Consumer Electronics (GCCE) · 2022
In order to improve sports skills, the comparison of the movement using the video is very effective for progress. Existing studies have been using posture similarity to map the motion timing. However, these methods have a disadvantage of a long execution time. Therefore, we propose a method to reduce the computational complexity by selectively using only data that is particularly effective among feature data used for mapping. Through evaluation experiments, we confirm that it is possible to reduce execution time without sacrificing performance even when the number of feature data is reduced. Based on the results of the evaluation experiment, we discuss the causes of the increase or decrease in accuracy for each motion and the factors that affect the accuracy of the mapping.