Action recognition using ensemble weighted multi-instance learning

By Guang Chen, Manuel Giuliani, Daniel Clarke, Andre Gaschler, Alois C. Knoll. 2014

23 citations recorded by the source corpus.

This paper deals with recognizing human actions in depth video data. Current state-of-the-art action recognition methods use hand-designed features, which are difficult to produce and time-consuming to extend to new modalities. In this paper, we propose a novel, 3.5D representation of a depth video for action recognition. A 3.5D graph of the depth video consists of a set of nodes that are the joints of the human body. Each joint is represented by a set of spatio-temporal features, which are computed by an unsupervised learning approach. However, if occlusions occur, the 3D positions of the joints are noisy which increases the intra-class variations in action classes. To address this problem, we propose the Ensemble Weighted Multi-Instance Learning approach (EnwMi) for the action recognition task. It considers the class imbalance and intra-class variations. We formulate the action recognition task with depth videos as a weighted multi-instance problem. We further integrate an ensemble learning method into the weighted multi-instance learning framework. Our approach is evaluated on Microsoft Research Action3D dataset, and the results show that it outperforms state-of-the-art methods.

Subfields: Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

Citations in the corpus

Space-Time Representation of People Based on 3D Skeletal Data: A Review (2016, neutral): “[189] EnwMi Learning BoW Lowlv Dict X X X Hussein et al.”
Bio-inspired predictive orientation decomposition of skeleton trajectories for real-time human activity prediction (2015, background): “Although many activity recognition methods [1] have been proposed in robotics applications, most of them focus on classification of finished activities [2], [3], [4].”
A discussion on the validation tests employed to compare human action recognition methods using the MSR Action3D dataset (2014, methodology): “ - 93.6 91.4 Super Normal Vector for Activity Recognition Using Depth Sequences, Yang and Tian [36] 2014 - - - 93.09? - Action Recognition Using Ensemble Weighted Multi-Instance Learning, Chen et al. [37] 2014 - - - 92? - Histogram of Oriented Displacements (HOD): Describing Trajectories of Human Jo”