arXivDaily arXiv每日学术速递 周一至周五更新
北美退伍军人节(Veterans Day),12 日无更新
arXiv 2012.14371cs.CV
收藏

用于动作识别的张量表示

Tensor Representations for Action Recognition

Data61/CSIRO(Data61/联邦科学与工业研究组织);Australian National University(澳大利亚国立大学);Mitsubishi Electric Research Labs (MERL)(三菱电机研究实验室)

Piotr Koniusz, Lei Wang, Anoop Cherian

摘要:

视频序列中的人类行为的特点是空间特征和它们的时间动态之间的复杂相互作用。在本文中,我们提出了新的张量表示复杂捕捉这种高阶视觉特征之间的关系,为任务的动作识别。我们提出了两种基于张量的特征表示,即(i)序列相容性核(SCK)和(ii)动力学相容性核(DCK)。SCK建立在特征之间的时空相关性上,而DCK明确地对序列的动作动态进行建模。我们还探索了SCK的泛化,创造了SCK(+),它对连续性进行操作以捕获相关性的局部-全局相互作用,它可以包含多模态输入,例如,骨架3D身体关节和每帧分类器得分,这些得分是从在视频上训练的深度学习模型获得的。我们引入线性化这些内核,导致紧凑和快速的描述符。我们提供实验(i)3D骨架动作序列,(ii)细粒度的视频序列,和(iii)标准的非细粒度的视频。由于我们的最终表示是捕获特征的高阶关系的张量,因此它们与用于鲁棒细粒度识别的同现相关。我们使用高阶张量和所谓的特征值功率归一化(EPN),长期以来一直推测执行高阶事件的谱检测,从而检测细粒度的特征关系,而不仅仅是计算动作序列中的特征。我们证明了一个张量的顺序r,建立从Z* 维的功能,再加上EPN确实检测到,如果至少有一个高阶的发生是'投影'到它的binom(Z*,r)的dim. r的子空间之一的张量表示,从而形成一个张量的功率归一化度量赋予binom(Z*,r)这样的'检测器'。

英文摘要:

Human actions in video sequences are characterized by the complex interplay between spatial features and their temporal dynamics. In this paper, we propose novel tensor representations for compactly capturing such higher-order relationships between visual features for the task of action recognition. We propose two tensor-based feature representations, viz. (i) sequence compatibility kernel (SCK) and (ii) dynamics compatibility kernel (DCK). SCK builds on the spatio-temporal correlations between features, whereas DCK explicitly models the action dynamics of a sequence. We also explore generalization of SCK, coined SCK(+), that operates on subsequences to capture the local-global interplay of correlations, which can incorporate multi-modal inputs e.g., skeleton 3D body-joints and per-frame classifier scores obtained from deep learning models trained on videos. We introduce linearization of these kernels that lead to compact and fast descriptors. We provide experiments on (i) 3D skeleton action sequences, (ii) fine-grained video sequences, and (iii) standard non-fine-grained videos. As our final representations are tensors that capture higher-order relationships of features, they relate to co-occurrences for robust fine-grained recognition. We use higher-order tensors and so-called Eigenvalue Power Normalization (EPN) which have been long speculated to perform spectral detection of higher-order occurrences, thus detecting fine-grained relationships of features rather than merely count features in action sequences. We prove that a tensor of order r, built from Z* dimensional features, coupled with EPN indeed detects if at least one higher-order occurrence is `projected' into one of its binom(Z*,r) subspaces of dim. r represented by the tensor, thus forming a Tensor Power Normalization metric endowed with binom(Z*,r) such `detectors'.

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。

↑