Action Recognition using Temporal Bag-of-Words from Depth Maps
Parul Shukla, Kanad Kishore Biswas, Prem Kumar Kalra · 2013
In this paper, we present a methodology for hu-man action recognition from a sequence of depth maps obtained using Microsoft Kinect. Specifically, we use a Temporal Bag-of-Words model as representa-tion scheme to capture the variation of features across the temporal domain. Our methodology builds the Temporal Bag-of-Words model on top of the spatio-temporal features extracted from interest points. The local spatio-temporal features provide some invariance to scale, viewpoint changes by capturing the local in-formation. In order to make the representation insen-sitive to temporal sequence misalignment, we propose using the Temporal Bag-of-Words model in a hierarchi-cal manner by recursively partitioning the depth maps sequence into sub-sequences in temporal domain. Clas-sification is done using SVM. We test our algorithm on our own dataset consisting of eight different actions. 1