Applying TS-DBN model into sports behavior recognition with deep learning approach

Abstract

The purposes are to automatically collect information about human sports behavior from massive video data and provide an explicit recognition and analysis of body movements. The analysis of multi-scale input data, the improvement of spatiotemporal Deep Belief Network (DBN), and the different pooling strategies are regarded as the focuses to improve the belief networks in deep learning (DL). Moreover, a human sports behavior recognition model is proposed based on particular spatio-temporal features. Also, video frame data are collected from the Royal Institute of Technology (KTH) and University of Central Florida (UCF) datasets for training. The TensorFlow platform is employed to simulate the built algorithm. Finally, the constructed algorithm model is compared with the DBN proposed by Yang et al. the Convolutional Neural Network (CNN) proposed by Ullah et al. and the DBN-Hidden Markov Model (HMM) algorithm proposed by Xu et al. to analyse its performance. The recognition effects of each algorithm in the two datasets are analyzed. Results demonstrate that CNN developed by Ullah et al. has the worst accuracy on the KTH and UCF datasets, followed by DBN developed by Yang et al. and DBN-HMM developed by Xu et al. The constructed algorithm model can provide the highest accuracy, reaching about 90%, and the recognition accuracy of human sports behaviors of each algorithm on the KTH dataset is lower than that on the UCF dataset. On the KTH dataset, the recognition accuracy for boxing is the highest and running the lowest. Analyzing the model’s accuracy in the four scenes (S1, S2, S3, and S4) on the KTH dataset suggests that the recognition accuracy for the indoor scene (S4) is higher than that of the outdoor scenes (S1, S2, and S3). On the UCF dataset, the recognition accuracy for lifting is the highest, reaching 99%, and that for walking is the lowest, reaching 51%. Therefore, the proposed human sports recognition model can provide a higher accuracy than other classic DL algorithms, providing an experimental basis for subsequent human sports recognition research.

Application of Distributed Probability Model in Sports Based on Deep Learning: Deep Belief Network (DL-DBN) Algorithm for Human Behaviour Analysis

“…, n, with such a total of n categories. Equation (1) shows the categorization probability presumed in soft max correlation classification for the test data c.…”

“…Significantly, the extraction of human activity knowledge from huge video data sources has become a pressing issue in a variety of fields. If intelligent video surveillance is used, video can also be instantly designed and analysed [1]. Human behaviours can also be accurately recognised in the real world and can generate security warnings with time settings.…”

Optimal Deep Convolutional Neural Network with Pose Estimation for Human Activity Recognition

“…The entire method is trained end-to-end for allowing effective representation to be created for the last action recognitions. In Guo and Wang [12], DBN is enhanced, also a human sport behaviour detection model based on specific spatio-temporal features are presented for obtaining, recognizing, and analyzing human sport behaviors from huge video data. The generated method is simulated on University of Central Florida (UCF) and Royal Institute of Technology (KTH) data sets, which provide an experiment for succeeding body detection and sports development in China.…”

Deep Neural Networks in Video Human Action Recognition: A review

“…The application scenario of DBN also includes sports recognition. The proposed [136] time-space deep belief network recognizes human sports behaviors, resulting in higher accuracy than CNN. Furthermore, contrastive divergence(CD) is the optimization algorithm of prediction accuracy using a weights matrix under DBN.…”