This study presents a new semi-supervised action recognition method via adaptive feature analysis. We assume that videos can be regarded as data points in embedding manifold subspace, and their matching problem quantified through specific Grassmannian kernel function while integrating correlation exploration similarity measurement into joint framework. By maximizing the intra-class compactness ...

The success of Zero-shot Action Recognition (ZSAR) methods is intrinsically related to the nature semantic side information used transfer knowledge, although this aspect has not been primarily investigated in literature. This work introduces a new ZSAR method based on relationships actions-objects and actions-descriptive sentences. We demonstrate that representing all object classes using descr...

Hand action recognition is a special case of with applications in human-robot interaction, virtual reality or life-logging systems. Building classifiers able to work for such heterogeneous domains very challenging. There are subtle changes across different actions from given application but also large variations (e.g. vs life-logging). This introduces novel skeleton-based hand motion representa...

We introduce a generative Bayesian switching dynamical model for action recognition in 3D skeletal data. Our encodes highly correlated data into few sets of low-dimensional temporal processes and from there decodes to the motion their associated labels. parameterize these with regard deep autoregressive prior accommodate both multimodal higher-order nonlinear inter-dependencies. This results la...

In this article, we propose a transformer-based RGB-D egocentric action recognition framework, called Trear. It consists of two modules: 1) interframe attention encoder and 2) mutual-attentional fusion block. Instead using optical flow or recurrent units, adopt self-attention mechanism to model the temporal structure data from different modalities. Input frames are cropped randomly mitigate eff...

It is well accepted that the rise in the proliferation of inexpensive digital media collection and manipulation devices has motivated the need to access this data by content rather than by keywords. The requirements of content based access are well understood by the digital media research community and there is no need to elaborate further here. Parsing multimedia streams by detection and class...

Automatically understanding human actions using motion trajectories derived from video sequences is a very challenging problem. Since an action takes place in 3-D, and is projected on 2-D image, depending on the viewpoint of the camera, the projected 2-D trajectory may vary. Therefore, the same action may have very different trajectories, and trajectories of different actions may look the same....

We propose a soft attention based model for the task of action recognition in videos. We use multi-layered Recurrent Neural Networks (RNNs) with Long Short-Term Memory (LSTM) units which are deep both spatially and temporally. Our model learns to focus selectively on parts of the video frames and classifies videos after taking a few glimpses. The model essentially learns which parts in the fram...

We introduce a simple yet surprisingly powerful model to incorporate attention in action recognition and human object interaction tasks. Our proposed attention module can be trained with or without extra supervision, and gives a sizable boost in accuracy while keeping the network size and computational cost nearly the same. It leads to significant improvements over state of the art base archite...