A structured multi-feature representation for recognizing human action and interaction

TL;DR: The proposed WHITE STAG model and kernel sliding perceptron outperformed the existing well known statistical state-of-the-art methods by achieving a weighted average recognition rate of 87.48% over UT-interaction, 87.5% over BIT-Interaction and 85.7% over IM-IntensityInteractive7 datasets.

TL;DR: A comprehensive review of human action and interaction recognition methods, covering both hand-crafted features and learning-based features, with a special focus on data captured by RGB-D sensors is provided.

TL;DR: Wang et al. as mentioned in this paper proposed a multimodal two-stream 3D network framework, which can exploit complementary multi-modal information to improve the recognition performance, and the classification scores from two streams are fused for action recognition.

TL;DR: The results show the benefits of using a hierarchical model that exploits the sharing and composition of body poses into atomic actions, and atomic actions into activities, and the compositional capabilities of the model bring robustness to body occlusions.

TL;DR: This work presents a residual learning framework to ease the training of networks that are substantially deeper than those used previously, and provides comprehensive empirical evidence showing that these residual networks are easier to optimize, and can gain accuracy from considerably increased depth.

TL;DR: This work investigates the effect of the convolutional network depth on its accuracy in the large-scale image recognition setting using an architecture with very small convolution filters, which shows that a significant improvement on the prior-art configurations can be achieved by pushing the depth to 16-19 weight layers.

TL;DR: In this paper, the authors investigated the effect of the convolutional network depth on its accuracy in the large-scale image recognition setting and showed that a significant improvement on the prior-art configurations can be achieved by pushing the depth to 16-19 layers.

TL;DR: Issues such as solving SVM optimization problems theoretical convergence multiclass classification probability estimates and parameter selection are discussed in detail.