Binary dense sift flow based two stream CNN for human action recognition

Park, Sang Kyoo; Chung, Jun Ho; Kang, Tae-Koo; Lim, Myo Taeg

doi:10.1007/s11042-021-10795-2

Cited by 13 publications

(18 citation statements)

References 45 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This section describes the proposed technique's comparison results, in which our novel technique is compared to baseline approaches such as volumetric Spatiograms of either the Local Binary Pattern (VS-LBP) [45], Local Binary Pattern (LBP) [46], Temporal Pyramid Matching of the Local Binary Pattern (TPM-LBP) [47], Pyramid Histogram of Gradients (PHOG) [48], as well as Scale Invariant Feature Transform (SIFT) [49].…”

Section: Comparison Resultsmentioning

confidence: 99%

“…Accuracy (%) VS-LBP [45] 92.7 LBP [46] 91.5 TPM-LBP [47] 96.5 PHOG [48] 94.6 SIFT [49] 97.6 Proposed Method 99.5…”

Section: Methodsmentioning

confidence: 99%

“…The suggested approach achieves improved accuracy by using an optimized capsule neural network. Our proposed approach outperformed the baseline Volumetric Spatiograms of Local Binary Pattern (VS-LBP) [45], Local Binary Pattern (LBP) [46], Temporal Pyramid Matching of Local Binary Pattern (TPM-LBP) [47], Pyramid Histogram of Gradients (PHOG) [47], as well as Scale Invariant Feature Transform (SIFT) [49] by 92.7, 91.5, 96.5, 94.6, and 97.6% and tabulated in Table 2. As a result, our unique approach has a greater accuracy of 99.5% than existing techniques.…”

Section: Methodsmentioning

confidence: 99%

See 2 more Smart Citations

Hand Gesture Recognizing Model Using Optimized Capsule Neural Network

Suni¹,

Gopakumar²

2022

View full text Add to dashboard Cite

Hand gestures are a sort of nonverbal communication that may be utilized for many diverse purposes, including deaf-mute interaction, robotic manipulation, human-computer interface (HCI), residential management, and healthcare usage. Moreover, most current research uses the artificial intelligence approach effectively to extract dense features from hand gestures. Since most of them used neural network models, the performance of the models influences the modification of the hyperparameter to enhance recognition accuracy. Therefore, our research proposed a capsule neural network, in which the internal computations on the inputs are better encapsulated by transforming the findings into a tiny vector of information outputs. Moreover, to increase the accuracy of recognizing hand gestures, the neural network has been optimized by inserting additional SoftMax layers before the output layer of the CapsNet. Subsequently, the findings of the tests were assessed and then compared. This developed approach has been beneficial across all tests when contrasted against state-of-the-art systems.

show abstract

Section: Comparison Resultsmentioning

confidence: 99%

“…Accuracy (%) VS-LBP [45] 92.7 LBP [46] 91.5 TPM-LBP [47] 96.5 PHOG [48] 94.6 SIFT [49] 97.6 Proposed Method 99.5…”

Section: Methodsmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Hand Gesture Recognizing Model Using Optimized Capsule Neural Network

Suni¹,

Gopakumar²

2022

View full text Add to dashboard Cite

show abstract

“…e convolution layer is responsible for extracting the local features of the input image to obtain different feature maps; the pooling layer reduces the dimension of the extracted features of the convolution layer to retain important information while reducing the risk of overfitting due to nonessential information. Common pooling layer settings include average pooling and maximum pooling; the full connection layer plays a classification role in the network and enables sample data classification by mapping the learned feature data to the space of sample markers [6][7][8][9][10][11].…”

Section: Network Profilementioning

confidence: 99%

The Motor Action Analysis Based on Deep Learning

Zhang

2022

Scientific Programming

View full text Add to dashboard Cite

For the slow speed and low accuracy of slow motor action recognition methods, this study proposes a motor action analysis method based on the CNN network and the softmax classification model. First, in order to obtain motor action feature information, by using static spatial features of BN-inception based on CNN network extracted actions and high-dimensional features of 3D ConvNet, then based on softmax classifier structure and realizing taxonomic recognition of the motor actions. Finally, through the decision-layer fusion and time semantic continuity optimization strategy, the motion action recognition accuracy is further improved and the more efficient motion action classification recognition is realized. The results show that the proposed method can complete the motor action analysis and achieve the classification recognition accuracy to 83.11%, which has certain practical value.

show abstract

“…CNN has advantages in terms of computing speed and can form a good balance between accuracy and effectiveness in short videos. In order to take advantage of CNN's strengths and make up for its shortcomings, CNN has progressed from the previous 2D-CNN and 3D-CNN to the current two-stream CNN [7][8][9][10]. This has achieved good results.…”

Section: Introductionmentioning

confidence: 99%

Optimization of Action Recognition Model Based on Multi-Task Learning and Boundary Gradient

Zhou

Wang

et al. 2021

Electronics

View full text Add to dashboard Cite

Recently, people’s demand for action recognition has extended from the initial high classification accuracy to the high accuracy of the temporal action detection. It is challenging to meet the two requirements simultaneously. The key to behavior recognition lies in the quantity and quality of the extracted features. In this paper, a two-stream convolutional network is used. A three-dimensional convolutional neural network (3D-CNN) is used to extract spatiotemporal features from the consecutive frames. A two-dimensional convolutional neural network (2D-CNN) is used to extract spatial features from the key-frames. The integration of the two networks is excellent for improving the model’s accuracy and can complete the task of distinguishing the start–stop frame. In this paper, a multi-scale feature extraction method is presented to extract more abundant feature information. At the same time, a multi-task learning model is introduced. It can further improve the accuracy of classification via sharing the data between multiple tasks. The experimental result shows that the accuracy of the modified model is improved by 10%. Meanwhile, we propose the confidence gradient, which can optimize the distinguishing method of the start–stop frame to improve the temporal action detection accuracy. The experimental result shows that the accuracy has been enhanced by 11%.

show abstract

Binary dense sift flow based two stream CNN for human action recognition

Cited by 13 publications

References 45 publications

Hand Gesture Recognizing Model Using Optimized Capsule Neural Network

Hand Gesture Recognizing Model Using Optimized Capsule Neural Network

The Motor Action Analysis Based on Deep Learning

Optimization of Action Recognition Model Based on Multi-Task Learning and Boundary Gradient

Contact Info

Product

Resources

About