GolfMate: Enhanced Golf Swing Analysis Tool through Pose Refinement Network and Explainable Golf Swing Embedding for Self-Training

Article Education & Educational Research

Dancing on the inside: A qualitative study on online dance learning with teacher-AI cooperation

Jiwon Kang et al.

Summary: Recent technologies have expanded opportunities for online dance learning, but student-teacher interaction remains challenging in a distant and asynchronous learning environment. To address this issue, DancingInside is introduced as an online dance learning system that utilizes an AI tutor to provide timely feedback and improve performance. A user study with 11 students and 4 teachers demonstrates the effectiveness of the AI tutor in supporting reflection and complementing human teacher feedback.

EDUCATION AND INFORMATION TECHNOLOGIES (2023)

添加到收藏夹

Proceedings Paper Computer Science, Cybernetics

AI Coach: A Motor Skill Training System using Motion Discrepancy Detection

Chen-Chieh Liao et al.

Summary: The study explores the use of neural networks to learn spatial and temporal features of advanced players in sports and detect fine-grained differences between motions. An AI Coach prototype application is developed to find motion differences and provide recommended motions for users to correct their forms. The study demonstrates that the system helps users understand differences and highlights the importance of coaching beginners in the early learning phases.

PROCEEDINGS OF THE 4TH AUGMENTED HUMANS INTERNATIONAL CONFERENCE 2023, AHS2023 (2023)

添加到收藏夹

Article Computer Science, Information Systems

A Lightweight Top-Down Multi-Person Pose Estimation Method Based on Symmetric Transformation and Global Matching

Yangyang Li et al.

Summary: This paper proposes a lightweight multi-person pose estimation method that solves the problems of target detection, human detection in crowded state, and model complexity through symmetric transformation and global matching, and experiments confirm its effectiveness and competitiveness.

IEEE ACCESS (2022)

添加到收藏夹

Article Computer Science, Information Systems

AI Golf: Golf Swing Analysis Tool for Self-Training

Chen-Chieh Liao et al.

Summary: This study proposes a golf swing analysis tool using neural networks to help beginners improve their sports skills by comparing motion sequences and specifying significant differences. The system provides interpretable clues and intermediate pose restoration to facilitate users' understanding and self-training. The unsupervised learning method used in this approach is significant for future applications and research in sports and skill training processes.

IEEE ACCESS (2022)

添加到收藏夹

Article Robotics

Pose Refinement Graph Convolutional Network for Skeleton-Based Action Recognition

Shijie Li et al.

Summary: This study introduces a highly efficient graph convolutional network that addresses the limitations of previous works by gradually fusing motion and spatial information through a parallel structure and reducing the temporal resolution early on. By refining poses before processing and explicitly addressing potential errors, the network achieves a better balance between accuracy, memory usage, and processing time.

IEEE ROBOTICS AND AUTOMATION LETTERS (2021)

添加到收藏夹

Article Computer Science, Artificial Intelligence

OpenPose: Realtime Multi-Person 2D Pose Estimation Using Part Affinity Fields

Zhe Cao et al.

Summary: This work presents a real-time approach to detect the 2D pose of multiple people in an image, achieving high accuracy and real-time performance regardless of the number of people in the image. By refining PAFs separately, both runtime performance and accuracy are significantly improved. Additionally, the first combined body and foot keypoint detector reduces inference time and maintains accuracy compared to running them sequentially.

IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE (2021)

添加到收藏夹

Article Computer Science, Information Systems