Text-controlled Motion Mamba: Text-Instructed Temporal Grounding of Human Motion

arXiv:2404.11375v1 Announce Type: new
Abstract: Human motion understanding is a fundamental task with diverse practical applications, facilitated by the availability of large-scale motion capture datasets. Recent studies focus on text-motion tasks, such as text-based motion generation, editing and question answering. In this study, we introduce the novel task of text-based human motion grounding (THMG), aimed at precisely localizing temporal segments corresponding to given textual descriptions within untrimmed motion sequences. Capturing global temporal information is crucial for the THMG task. However, …

abstract applications arxiv availability cs.cv cs.mm datasets diverse editing focus human mamba motion capture novel practical question question answering scale studies study tasks temporal text type understanding