STMixer: A One-Stage Sparse Action Detector

arXiv:2404.09842v1 Announce Type: new
Abstract: Traditional video action detectors typically adopt the two-stage pipeline, where a person detector is first employed to generate actor boxes and then 3D RoIAlign is used to extract actor-specific features for classification. This detection paradigm requires multi-stage training and inference, and the feature sampling is constrained inside the box, failing to effectively leverage richer context information outside. Recently, a few query-based action detectors have been proposed to predict action instances in an end-to-end manner. However, …

abstract actor arxiv box classification cs.cv detection detectors extract feature features generate inference inside paradigm person pipeline sampling stage training type video