arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1805.04026cs.CV
收藏

走向行动的明确表示

Towards an Unequivocal Representation of Actions

University of Bristol(布里斯托尔大学)

Michael Wray, Davide Moltisanti, Dima Damen

摘要:

这项工作介绍了动词只表示动作和交互;描述类似的运动(例如“开门”,“打开橱柜”)的问题,并区分不同的(例如“开门”与“开瓶”)使用动词只标签。当前的动作识别方法忽略了动词之间合理的语义歧义和类重叠(图1),依赖于对象来消除交互的歧义。我们偏离了单动词标签,并引入了观察和多个动词标签之间的映射-为了创建一个明确的动作表示。新的表示受益于增加的词汇量和软分配到丰富的动词标签空间。我们使用双流融合CNN将这些表示作为多输出回归进行学习。所提出的方法优于传统的单动词标签(也称为多数表决)的三个自我中心的数据集的识别和检索。

英文摘要:

This work introduces verb-only representations for actions and interactions; the problem of describing similar motions (e.g. 'open door', 'open cupboard'), and distinguish differing ones (e.g. 'open door' vs 'open bottle') using verb-only labels. Current approaches for action recognition neglect legitimate semantic ambiguities and class overlaps between verbs (Fig. 1), relying on the objects to disambiguate interactions. We deviate from single-verb labels and introduce a mapping between observations and multiple verb labels - in order to create an Unequivocal Representation of Actions. The new representation benefits from increased vocabulary and a soft assignment to an enriched space of verb labels. We learn these representations as multi-output regression, using a two-stream fusion CNN. The proposed approach outperforms conventional single-verb labels (also known as majority voting) on three egocentric datasets for both recognition and retrieval.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑