A hardware friendly algorithm for action recognition using spatio-temporal motion-field patches

Bao, Ruihan; Shibata, Tadashi

doi:10.1016/j.neucom.2011.12.041

Cited by 10 publications

(3 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The forward kinematics-based reorientation algorithm uses the root joint point position and given parameters such as quaternion and model bone length to sequentially calculate the position of each joint point in the model under the model coordinate system [30]. The specific process is as follows: the position of the root joint point is updated with the initial transformation matrix, and when the position of the root joint point is determined, the model coordinate system is then determined.…”

Section: Action Redirection Algorithmmentioning

confidence: 99%

An analysis of the application of computer-assisted instruction in the teaching of physical education yoga

Qian

2023

Applied Mathematics and Nonlinear Sciences

View full text Add to dashboard Cite

In this paper, according to the TOF method in the principle of Kinect V2.0 depth measurement, the Kinect data stream can be obtained, which can realize the real-time acquisition of the user’s yoga movement data. On the basis of the extraction of joint motion features, the position of each joint point in the model under the model coordinate system is calculated sequentially using the action redirection algorithm. For the intermittent problem of keyframes in the yoga action screen of the joints, it is necessary to optimize the analysis of yoga action recognition through the action keyframe interpolation method. Starting from the demand for sports yoga teaching, we determine the design scheme of a computer-aided yoga teaching system, realize the design of a computer-aided yoga teaching system based on the system development platform and programming program, and test the performance of the computer-aided yoga teaching system by using simulation analysis. The results show that the yoga movement redirection algorithm, when matching the data information of the same movement, the distance value is 1 to 2 orders of magnitude smaller compared to the matching distance between different yoga movement sequences, and the use of the joint angle value can find out the joints where the trainer’s movement is not up to the standard, and the yoga trainer can reasonably adjust the movement based on the similarity and the joint angle value given by the system to achieve the purpose of yoga training.

show abstract

Section: Action Redirection Algorithmmentioning

confidence: 99%

An analysis of the application of computer-assisted instruction in the teaching of physical education yoga

Qian

2023

Applied Mathematics and Nonlinear Sciences

View full text Add to dashboard Cite

show abstract

“…In particular, the K-means algorithm has been employed in intelligent image processing applications such as object extraction 28) and action recognition. 29) In these applications, real-time performance is mandatory. However, since it requires a number of distance calculations and iterations, real time response is very difficult to achieve in software solutions.…”

Section: Introductionmentioning

confidence: 99%

“…30, outperformed a 2 GHz CPU by 20,000-fold and the OpenCV library (including much optimization for modern processors) running on Core2 Duo CPU (3 GHz) by 14,000-fold. Such a speed is essential for real-time applications such as those for image processing, 18,28,29) and would presumably play an essential role in space rovers. 16) The organization of this paper is as follows.…”

Section: Introductionmentioning

confidence: 99%

Real-Time Very Large-Scale Integration Recognition System with an On-Chip Adaptive K-Means Learning Algorithm

Hou

Zhu

et al. 2013

Jpn. J. Appl. Phys.

Self Cite

View full text Add to dashboard Cite

A very large-scale integration (VLSI) recognition system equipped with an on-chip learning capability has been developed for real-time processing applications. This system can work in two functional modes of operation: adaptive K-means learning mode and recognition mode. In the adaptive K-means learning mode, the variance ratio criterion (VRC) has been employed to evaluate the quality of K-means classification results, and the evaluation algorithm has been implemented on the chip. As a result, it has become possible for the system to autonomously determine the optimum number of clusters (K). In the recognition mode, the nearest-neighbor search algorithm is very efficiently carried out by the fully parallel architecture employed in the chip. In both modes of operation, many hardware resources are shared and the functionality is flexibly altered by the system controller designed as a finite-state machine (FSM). The chip is implemented on Altera Cyclone II FPGA with 46K logic cells. Its operating clock is 25 MHz and the processing times for adaptive learning and recognition with 256 64-dimension feature vectors are about 0.42 ms and 4 µs, respectively. Both adaptive K-means learning and recognition functions have been verified by experiments using the image data from the COIL-100 (Columbia University Object Image Library) database.

show abstract

Probabilistic spatio-temporal inference for motion event understanding

Choi

Lee

et al. 2013

Neurocomputing

View full text Add to dashboard Cite

A hardware friendly algorithm for action recognition using spatio-temporal motion-field patches

Cited by 10 publications

References 23 publications

An analysis of the application of computer-assisted instruction in the teaching of physical education yoga

An analysis of the application of computer-assisted instruction in the teaching of physical education yoga

Real-Time Very Large-Scale Integration Recognition System with an On-Chip Adaptive K-Means Learning Algorithm

Probabilistic spatio-temporal inference for motion event understanding

Contact Info

Product

Resources

About