Definition and Composition of Motor Primitives Using Latent Force Models and Hidden Markov Models

详细信息    查看全文

• 关键词：Movement representation ; Motor primitives ; Latent force models ; Hidden Markov models ; Switched models ; Multi ; output GPs
• 刊名：Lecture Notes in Computer Science
• 出版年：2017
• 出版时间：2017
• 年：2017
• 卷：10125
• 期：1
• 页码：249-256
• 丛书名：Progress in Pattern Recognition, Image Analysis, Computer Vision, and Applications
• ISBN：978-3-319-52277-7
• 卷排序：10125

文摘

In this work a different probabilistic motor primitive parameterization is proposed using latent force models (LFMs). The sequential composition of different motor primitives is also addressed using hidden Markov models (HMMs) which allows to capture the redundancy over dynamics by using a limited set of hidden primitives. The capability of the proposed model to learn and identify motor primitive occurrences over unseen movement realizations is validated using synthetic and motion capture data.