3D Pose-Based Temporal Action Segmentation for Figure Skating: A Fine-Grained and Jump Procedure-Aware Annotation Approach
This work addresses the lack of datasets and methods for 3D pose-based action segmentation in figure skating, which is incremental as it builds on existing TAS research with new data and annotations.
The study tackled the problem of temporal action segmentation in figure skating by creating the FS-Jump3D dataset using 3D pose data and a fine-grained annotation method, resulting in validated improvements for TAS models in this domain.
Understanding human actions from videos is essential in many domains, including sports. In figure skating, technical judgments are performed by watching skaters' 3D movements, and its part of the judging procedure can be regarded as a Temporal Action Segmentation (TAS) task. TAS tasks in figure skating that automatically assign temporal semantics to video are actively researched. However, there is a lack of datasets and effective methods for TAS tasks requiring 3D pose data. In this study, we first created the FS-Jump3D dataset of complex and dynamic figure skating jumps using optical markerless motion capture. We also propose a new fine-grained figure skating jump TAS dataset annotation method with which TAS models can learn jump procedures. In the experimental results, we validated the usefulness of 3D pose features as input and the fine-grained dataset for the TAS model in figure skating. FS-Jump3D Dataset is available at https://github.com/ryota-skating/FS-Jump3D.