| Autor: |
Sheng-Ru Hsiao, Chyuan-Huei Thomas Yang, Chien-Chang Chen, Cheng-Shian Lin |
| Předmět: |
|
| Zdroj: |
International Journal of Design, Analysis & Tools for Integrated Circuits & Systems; Feb2024, Vol. 13 Issue 1, p16-20, 6p |
| Abstrakt: |
Rowing is a sport that requires players to demonstrate high-density coordination and consistency in their movements. Stable and consistent rowing strokes by players lead to better performance. This study analyzes rowing movements recorded from rowing machine to measure the stability of single player. The proposed method first uses the OpenPose system to extract body joint poinst from rowing video. The joint point information was then fed into the Temporal Cycle Consistency (TCC) learning network, named as Pose Temporal Cycle-Consistency (PTCC), for training the model. The proposed PTCC network aligned the timelines of the videos as closely as possible and generated frame-by-frame retrieval results. At last, the average matched frames among each cycled rowing video represents the players rowing stability. Experimental results show that using some players' rowing videos can train the PTCC model and the similarity calculated from average matched frame rates is efficient to determine the stability of one player. [ABSTRACT FROM AUTHOR] |
| Databáze: |
Complementary Index |
| Externí odkaz: |
|
