Motion Style Extraction Based on Sparse Coding Decomposition
Xuan Thanh Nguyen, Thanh Ha Le, Hongchuan Yu

TL;DR
This paper introduces a sparse coding framework for decomposing and synthesizing motion styles, utilizing dynamic time warping for synchronization and limb length constraints for realism, enabling efficient and natural motion generation.
Contribution
It proposes a novel sparse coding-based approach for motion style decomposition and synthesis that requires less time and no manual alignment.
Findings
Generated smooth and natural motions
Reduced processing time compared to existing methods
Effective motion transfer with minimal manual intervention
Abstract
We present a sparse coding-based framework for motion style decomposition and synthesis. Dynamic Time Warping is firstly used to synchronized input motions in the time domain as a pre-processing step. A sparse coding-based decomposition has been proposed, we also introduce the idea of core component and basic motion. Decomposed motions are then combined, transfer to synthesize new motions. Lastly, we develop limb length constraint as a post-processing step to remove distortion skeletons. Our framework has the advantage of less time-consuming, no manual alignment and large dataset requirement. As a result, our experiments show smooth and natural synthesized motion.
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsHuman Motion and Animation · Video Analysis and Summarization · Human Pose and Action Recognition
