Searching for variable-speed motions in long sequences of motion capture data

作者：

Highlights：

• Subsequence matching algorithm for motion capture data that searches for variable-speed motions.

• A unique combination of elastic similarity measure and multi-level segmentation structure.

• Real-time searching in months of motion capture recordings by using a bit-vector representation or disk-based index.

• Effective and indexable 4096-dimensional motion features extracted using a neural network.

摘要

•Subsequence matching algorithm for motion capture data that searches for variable-speed motions.•A unique combination of elastic similarity measure and multi-level segmentation structure.•Real-time searching in months of motion capture recordings by using a bit-vector representation or disk-based index.•Effective and indexable 4096-dimensional motion features extracted using a neural network.

论文关键词：Content-based retrieval,Motion capture data,Subsequence matching,Speed-invariant retrieval,Similarity measure,Hierarchical segmentation,Indexing

论文评审过程：Received 10 March 2017, Revised 26 October 2017, Accepted 11 April 2018, Available online 12 April 2018, Version of Record 6 December 2018.

论文官网地址：https://doi.org/10.1016/j.is.2018.04.002