Motion Trajectory for Human Action Recognition Using Fourier Temporal Features of Skeleton Joints

Kumar, Naresh; Sukavanam, N.

doi:10.18178/joig.6.2.174-180

Cited by 42 publications

(6 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Although the three-dimensional triangulation method works well, it does not mean that machines must “see” the world in the same way as humans. To make a method achieve robust results, different sensing technologies can be considered [ 16 , 17 ]. Different from the stereo triangulation method, camera using time-of-flight method (TOF) emits light pulses directly from a single camera, and the light is reflected back to the camera by the objects in the scene.…”

Section: Human Motion Recognition Algorithm Based On Depth Informationmentioning

confidence: 99%

Design of Children’s Motor Skills Training Management System Based on Human Movement Recognition Algorithm of Depth Information

Cheng

2022

Computational Intelligence and Neuroscience

View full text Add to dashboard Cite

With the deepening of the concept of all-round development, the training of children’s motor skills is becoming more and more popular. Children’s motor skills training management system is based on the product of children’s motor skills training combined with modern technology. This article aims to design a management system for children’s motor skills training based on human motion recognition methods. This article proposes a human motion recognition algorithm based on depth information. The traditional human motion recognition algorithm is more biased towards children’s motor skills. For the design of the system, this article also pays great attention to children’s experience. After analyzing the performance of the system in this study, it is found that the system is not ideal for multitarget human motion recognition. Therefore, this article optimizes the recognition. After the optimization, the recognition rate of the head, upper body, and lower body of the multiobjective human motion recognition is more than 80%.

show abstract

Section: Human Motion Recognition Algorithm Based On Depth Informationmentioning

confidence: 99%

Design of Children’s Motor Skills Training Management System Based on Human Movement Recognition Algorithm of Depth Information

Cheng

2022

Computational Intelligence and Neuroscience

View full text Add to dashboard Cite

show abstract

“…Recently, action recognition in simple scenes has gained much attention [5]. The Two-Stream network [8] takes RGB frames and dense optical flow as the input of spatial and temporal streams to learn action features in different dimensions.…”

Section: Action Recognitionmentioning

confidence: 99%

Basketball action recognition algorithm based on global context awareness

Luo

2023

Fifteenth International Conference on Machine Vision (ICMV 2022)

View full text Add to dashboard Cite

Basketball action recognition is the basis of intellectual analysis of basketball videos. Different from action recognition in simple scenes, the difference between classes of basketball actions is slight, and the backgrounds in the video are very similar. Therefore, it is not easy to recognize the basketball actions directly based on short-term temporal information or the scene information in the video. A Global Context-Aware Network (GCA-Net) for basketball action recognition is proposed to address this problem in this paper. It contains a Multi-Time Scale Aggregation (MTSA) module and a Spatial-Channel Interaction (SCI) module to process multiple types of information on feature layers. The MTSA module uses a temporal pyramid to get contextual links in the temporal dimension through one-dimensional convolution with different dilation rates. The SCI module enhances the feature representation to obtain more prosperous category attributes and spatial information by interacting with information across dimensions. We conducted experiments on the basketball action recognition dataset SpaceJam, and the results show that GCA-Net can effectively classify basketball actions. The average recognition accuracy of ten types of basketball actions in the dataset is 91.54%, which is an improvement compared with the current mainstream methods.

show abstract

“…Since this research, many two-stream CNN-based methods have been proposed to improve action recognition performance 12,13,14,15,16 . There are also human action recognition methods based on LSTM 17 and Skeleton Joints 18 . Due to the advent of large-scale video datasets, a recent popular approach is to employ 3D CNNs with convolutional kernels containing spatiotemporal information 8 .…”

Section: Related Work 21 Action Recognitionmentioning

confidence: 99%

SlowFast with DropBlock and smooth samples loss for student action recognition

Bao²,

Chen³

et al. 2022

Fourteenth International Conference on Digital Image Processing (ICDIP 2022)

View full text Add to dashboard Cite

Due to the advent of large-scale video datasets, action recognition using three-dimensional convolutions (3D CNNs) containing spatiotemporal information has become mainstream. Aiming at the problem of classroom student behavior recognition, the paper adopts the improved SlowFast network structure to deal with spatial structure and temporal events respectively. First, DropBlock (a regularization method) is added to the SlowFast network to solve the overfitting problem. Second, for the problem of Long-Tailed Distribution, the designed Smooth Sample (SS) Loss function is added to the network to smooth the number of samples. Classification experiments show that compared with similar methods, the model accuracy of our method on the Kinetics and Student Action Dataset is increased by 2.1% and 2.9%, respectively.

show abstract

Motion Trajectory for Human Action Recognition Using Fourier Temporal Features of Skeleton Joints

Cited by 42 publications

References 26 publications

Design of Children’s Motor Skills Training Management System Based on Human Movement Recognition Algorithm of Depth Information

Design of Children’s Motor Skills Training Management System Based on Human Movement Recognition Algorithm of Depth Information

Basketball action recognition algorithm based on global context awareness

SlowFast with DropBlock and smooth samples loss for student action recognition

Contact Info

Product

Resources

About