Temporal sparse feature auto‐combination deep network for video action recognition

Abstract

Summary In order to deal with action recognition for large‐scale video data, we present a spatio‐temporal auto‐combination deep network, which is able to extract deep features from short video segments by making full use of temporal contextual correlation of corresponding pixels among successive video frames. Based on conventional sparse encoding, we further consider the representative features in adjacent nodes of the hidden layers according to activation states similarities. A sparse auto‐combination strategy is applied to multiple input maps in each convolution stage. An information constraint of the representative features of hidden layer nodes is imposed to handle the adaptive sparse encoding of the topology. As a result, the learned features can represent the spatio‐temporal transition relationships better and the number of hidden nodes can be restricted to a certain range. We conduct a series of experiments on two public data sets. The experimental results show that our approach is more effective and robust in video action recognition compared with traditional methods.

uMoDT: an unobtrusive multi-occupant detection and tracking using robust Kalman filter for real-time activity recognition

“…In the following subsection, to show the generality of the TVS-AR method, we describe and evaluate the proposed CNN-based model using the T V S F seq for AR. We present the classification results to prove the performance and suitability of the presented approach using low-resolution T V S F seq in terms of accuracy [65]. We used frame-based approach for recognizing 16 different activities showing the efficacy of a model by demonstrating it for a high HAR accuracy score of approximately 90.99%.…”

Real‐time action feature extraction via fast PCA‐Flow

“…Action recognition, as an important research area of IoT, has a wide application prospect in daily scenes such as automatic driving, video surveillance, etc. Much works have been recently devoted to action recognition, and among them, local spatiotemporal features are shown to be successful on a variety of challenging action recognition datasets 1–3 …”

High performance deep learning techniques for big data analytics

“…CNNs have also been widely applied in video content analysis. Wang et al apply CNN networks for automatic recognition of human actions in surveillance videos.…”