Search

Video-based human action recognition is currently one of the most active research areas in computer vision. Various research studies indicate that the performance of action recognition is highly dependent on the type of features being extracted and how the actions are represented. Since the release of the Kinect camera, a large number of Kinect-based human action recognition techniques have been proposed in the literature. However, there still does not exist a thorough comparison of these Kinect-based techniques under the grouping of feature types, such as handcrafted versus deep learning features and depth-based versus skeleton-based features. In this paper, we analyze and compare 10 recent Kinect-based algorithms for both cross-subject action recognition and cross-view action recognition using six benchmark datasets. In addition, we have implemented and improved some of these techniques and included their variants in the comparison. Our experiments show that the majority of methods perform better on cross-subject action recognition than cross-view action recognition, that the skeleton-based features are more robust for cross-view recognition than the depth-based features, and that the deep learning features are suitable for large datasets. [ABSTRACT FROM AUTHOR]

Performance evaluation of multi-target tracking algorithms is of great practical importance in the design, parameter optimization and comparison of tracking systems. The goal of performance evaluation is to measure the distance between two sets of tracks: the ground truth tracks and the set of estimated tracks. This paper proposes a mathematically rigorous metric for this purpose. The basis of the proposed distance measure is the recently formulated consistent metric for performance evaluation of multi-target filters, referred to as the OSPA metric. Multi-target filters sequentially estimate the number of targets and their position in the state space. The OSPA metric is therefore defined on the space of finite sets of vectors. The distinction between filtering and tracking is that tracking algorithms output tracks and a track represents a labeled temporal sequence of state estimates, associated with the same target. The metric proposed in this paper is therefore defined on the space of finite sets of tracks and incorporates the labeling error. Numerical examples demonstrate that the proposed metric behaves in a manner consistent with our expectations. [ABSTRACT FROM PUBLISHER]