3D-TDC: A 3D temporal dilation convolution framework for video action recognition

Ming, Yue; Feng, Fan; Li, Chao; Xue, Jing-Hao

doi:10.1016/j.neucom.2021.03.120

Cited by 21 publications

(4 citation statements)

References 52 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…+e methods based on traditional handcrafted features have relatively low accuracy, poor robustness, and universality and are suitable for relatively simple gesture and action recognition with low hardware requirements but are difficult to handle human behavior recognition in complex scenes. +erefore, in general, the traditional method of feature extraction has poor generalization and is complicated to implement [12][13][14][15][16].…”

Section: Introductionmentioning

confidence: 99%

[Retracted] Video Tactical Intelligence Analysis Method of Karate Competition Based on Convolutional Neural Network

Zhong

2022

Discrete Dynamics in Nature and Society

View full text Add to dashboard Cite

The performance of image classification technology based on deep network has been greatly improved, making computer vision enter the stage of industrialization and be gradually applied to many aspects of human work and life. As a typical classification task in computer vision, human behavior recognition has immeasurable potential value in medical, family, transportation, and other scenarios. At the same time, in the field of competitive sports, the integration of artificial intelligence technology and sports technical and tactical analysis is undoubtedly an important way to innovate and improve the technical and tactical level. Taking karate as an example, the study of athletes’ training and competition videos is an important means and method for technical and tactical analysis in competitive sports. Traditional tactical intelligence analysis methods have many shortcomings, such as high labor cost, serious data loss, long delay, and low accuracy. Therefore, based on the convolutional neural network, this paper establishes a new graph convolution model for automatic intelligent analysis of karate athletes’ technical action recognition, action frequency statistics, and trajectory tracking. The technology effectively makes up for the disadvantages of traditional tactical intelligence analysis methods. The research results show that the new topology map construction method has a significant effect on improving the accuracy of behavior recognition and also lays a foundation for technical and tactical analysis.

show abstract

Section: Introductionmentioning

confidence: 99%

[Retracted] Video Tactical Intelligence Analysis Method of Karate Competition Based on Convolutional Neural Network

Zhong

2022

Discrete Dynamics in Nature and Society

View full text Add to dashboard Cite

show abstract

“…The 3D CNN-based framework has spatiotemporal modeling capabilities and improves the performance of video action recognition models. 3D ConvNets [31,3,11,24] extended 2D image models [14,36] to the spatial-temporal domain, treating spatial and temporal dimensions in the same way. C3D [31] stacked spatiotemporal convolution kernels to efficiently represent video dense structure.…”

Section: D Cnnmentioning

confidence: 99%

Cross-Enhancement Transformer for Action Segmentation

Wang¹,

Wang²,

Zhuang³

et al. 2022

Preprint

View full text Add to dashboard Cite

Temporal convolutions have been the paradigm of choice in action segmentation, which enhances long-term receptive fields by increasing convolution layers. However, high layers cause the loss of local information necessary for frame recognition. To solve the above problem, a novel encoder-decoder structure is proposed in this paper, called Cross-Enhancement Transformer. Our approach can be effective learning of temporal structure representation with interactive self-attention mechanism. Concatenated each layer convolutional feature maps in encoder with a set of features in decoder produced via self-attention. Therefore, local and global information are used in a series of frame actions simultaneously. In addition, a new loss function is proposed to enhance the training process that penalizes over-segmentation errors. Experiments show that our framework performs state-of-the-art on three challenging datasets: 50Salads, Georgia Tech Egocentric Activities and the Breakfast dataset.

show abstract

“…3-D CNN directly accepts continuous video frames as input. It uses a 3-D convolution kernel to extract spatial and temporal features simultaneously through series convolution, which shows greater advantages than 2-D CNN in the field of video processing [29], [30]. The 3-D CNN was first proposed by Ji et al [31] and applied to human behavior recognition.…”

Section: B Video Detection Of Pipeline Leakagementioning

confidence: 99%

High Reliability Pipeline Leakage Detection Based on Machine Vision in Complex Industrial Environment

Lyu

Zhang

et al. 2022

IEEE Sensors J.

View full text Add to dashboard Cite

The vision-based pipeline leakage detection is an intelligent leak detection method based on the industrial Internet of Things monitoring platform. It has the advantages of high safety factor and detection visualization. However, in the actual complex industrial environment, there are some problems, such as environmental interference and transmission quality of sensing networks. These low-quality sensing image data bring noise to the vision-based pipeline leakage detection, resulting in the risk of missed detection and false detection. In view of the above problems, we propose a highly reliable pipeline leakage detection method based on machine vision in the complex industrial environment. First, we propose a key frame selection method based on a lightweight image quality assessment, which can adapt to the complex detection environment. The key frame containing available feature information is selected to eliminate the interference of low-quality sensing image data on subsequent feature extraction and realize pipeline leakage video denoising. Then, the C3D network is used to extract space-time features at the same time to detect the leakage of pipeline leakage video. The experimental results show that the effect of our proposed method is better than other existing methods in the complex industrial environment. When the noisy data ratio of the detected image is 15%, the accuracy can be improved by 2.6 percentage points, up to 97.8%, which ensures the reliability of the pipeline leakage detection in the complex industrial environment.

show abstract

3D-TDC: A 3D temporal dilation convolution framework for video action recognition

Cited by 21 publications

References 52 publications

[Retracted] Video Tactical Intelligence Analysis Method of Karate Competition Based on Convolutional Neural Network

[Retracted] Video Tactical Intelligence Analysis Method of Karate Competition Based on Convolutional Neural Network

Cross-Enhancement Transformer for Action Segmentation

High Reliability Pipeline Leakage Detection Based on Machine Vision in Complex Industrial Environment

Contact Info

Product

Resources

About