Learning Compact Features for Human Activity Recognition Via Probabilistic First-Take-All
Learning Compact Features for Human Activity Recognition Via Probabilistic First-Take-All
复制标题
DOI：
10.1109/tpami.2018.2874455
复制
发表时间：
2020-01-01
影响因子：
23.6
通讯作者：
Hua, Kien A.
中科院分区：
文献类型：
--
作者：
Ye, Jun;Qi, Guo-Jun;Hua, Kien A.
随着移动传感器技术的普及，智能可穿戴设备通过从多维日常传感器信号中学习富有表现力的表征，为解决具有挑战性的人类活动识别(HAR)问题提供了前所未有的机会。这启发了我们开发一种新的算法，既适用于基于摄像头的HAR系统，也适用于基于可穿戴传感器的HAR系统。虽然已经报道了竞争性的分类精度，但现有的方法经常面临着区分由不同时间顺序的活动模式组成的视觉相似活动的挑战。在本文中，我们提出了一种新的概率算法来紧凑地编码HAR的活动模式的时间顺序。具体地说，该算法学习一组最优的潜在模式，以便它们的时间结构在识别不同的人类活动时真正起到作用。然后，引入一种新的概率优先通吃(PFTA)方法从这些潜在模式的阶数中生成紧致特征来对整个序列进行编码，并通过紧致特征之间的汉明距离来有效地度量不同序列之间的时间结构相似性。在三个公开的HAR数据集上的实验表明，所提出的pFTA方法在准确率和效率方面都能获得与之相当的性能。
With the popularity of mobile sensor technology, smart wearable devices open a unprecedented opportunity to solve the challenging human activity recognition (HAR) problem by learning expressive representations from the multi-dimensional daily sensor signals. This inspires us to develop a new algorithm applicable to both camera-based and wearable sensor-based HAR systems. Although competitive classification accuracy has been reported, existing methods often face the challenge of distinguishing visually similar activities composed of activity patterns in different temporal orders. In this paper, we propose a novel probabilistic algorithm to compactly encode temporal orders of activity patterns for HAR. Specifically, the algorithm learns an optimal set of latent patterns such that their temporal structures really matter in recognizing different human activities. Then, a novel probabilistic First-Take-All (pFTA) approach is introduced to generate compact features from the orders of these latent patterns to encode the entire sequence, and the temporal structural similarity between different sequences can be efficiently measured by the Hamming distance between compact features. Experiments on three public HAR datasets show the proposed pFTA approach can achieve competitive performance in terms of accuracy as well as efficiency.