A Cuboid CNN Model With an Attention Mechanism for Skeleton-Based Action Recognition

Zhu, Kaijun; Wang, Ruxin; Zhao, Qingsong; Cheng, Jun; Tao, Dapeng

doi:10.1109/tmm.2019.2962304

Cited by 55 publications

(23 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Gradually, this method permeates the research of computer vision. The self-attention mechanism [22,23,24] allows the neural network to selectively focus on a particular part of the input and extract useful neural network perception information. Vaswani's groundbreaking work [25] proposed the selfattention mechanism for the first time and successfully applied it to natural language processing solutions.…”

Section: B Self-attentionmentioning

confidence: 99%

Temporal Transformer Networks with Self-Supervision for Action Recognition

Zhang¹,

Li²,

Wu³

et al. 2021

Preprint

View full text Add to dashboard Cite

In recent years, 2D Convolutional Networksbased video action recognition has encouragingly gained wide popularity; However, constrained by the lack of long-range non-linear temporal relation modeling and reverse motion information modeling, the performance of existing models is, therefore, undercut seriously. To address this urgent problem, we introduce a startling Temporal Transformer Network with Self-supervision (TTSN). Our high-performance TTSN mainly consists of a temporal transformer module and a temporal sequence self-supervision module. Concisely speaking, we utilize the efficient temporal transformer module to model the nonlinear temporal dependencies among non-local frames, which significantly enhances complex motion feature representations. The temporal sequence self-supervision module we employ unprecedentedly adopts the streamlined strategy of "random batch random channel" to reverse the sequence of video frames, allowing robust extractions of motion information representation from inversed temporal dimensions and improving the generalization capability of the model. Extensive experiments on three widely used datasets (HMDB51, UCF101, and Something-something V1) have conclusively demonstrated that our proposed TTSN is promising as it successfully achieves state-of-the-art performance for action recognition.

show abstract

Section: B Self-attentionmentioning

confidence: 99%

Temporal Transformer Networks with Self-Supervision for Action Recognition

Zhang¹,

Li²,

Wu³

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…This work was supported in part by the Natural Sciences and Engineering Research Council of Canada through a Discovery Grant. signal processing, including, among others, image dehazing [22]- [24], salient object detection [25], [26], action recognition [27], [28], and video captioning [29], [30]. In particular, it enables the development of data-driven approaches to VFI and super-resolution (SR) [31]- [34] that can capitalize on the learning capability of neural networks as opposed to relying on prescribed rules.…”

Section: Introductionmentioning

confidence: 99%

Learning for Unconstrained Space-Time Video Super-Resolution

Shi

Xiao-hong

et al. 2022

IEEE Trans. on Broadcast.

View full text Add to dashboard Cite

“…The depth image provides range information of the object of interest based on which a 3D reconstruction can be generated. Commodity depth cameras have been used widely in various applications [5][6] [7], among which several scanning systems were proposed for body shape reconstruction [8][9] [10]. 3D body shape reconstruction methods can be classified mainly into three categories: non-parametric, parametric, and template-based methods.…”

Section: Introductionmentioning

confidence: 99%

3DBodyNet: Fast Reconstruction of 3D Animatable Human Body Shape From a Single Commodity Depth Camera

Munteanu

2022

IEEE Trans. Multimedia

View full text Add to dashboard Cite

Knowledge about individual body shape has numerous applications in various domains such as healthcare, fashion and personalized entertainment. Most of the depth based whole body scanners need multiple cameras surrounding the user and requiring the user to keep a canonical pose strictly during capturing depth images. These scanning devices are expensive and need professional knowledge for operation. In order to make 3D scanning as easy-to-use and fast as possible, there is a great demand to simplify the process and to reduce the hardware requirements. In this paper, we propose a deep learning algorithm, dubbed 3DBodyNet, to rapidly reconstruct the 3D shape of human bodies using a single commodity depth camera. As easy-to-use as taking a photo using a mobile phone, our algorithm only needs two depth images of the front-facing and back-facing bodies. The proposed algorithm has strong operability since it is insensitive to the pose and the pose variations between the two depth images. It can also reconstruct an accurate body shape for users under tight/loose clothing. Another advantage of our method is the ability to generate an animatable human body model. Extensive experimental results show that the proposed method enables robust and easy-to-use animatable human body reconstruction, and outperforms the state-of-the-art methods with respect to running time and accuracy.

show abstract

A Cuboid CNN Model With an Attention Mechanism for Skeleton-Based Action Recognition

Cited by 55 publications

References 33 publications

Temporal Transformer Networks with Self-Supervision for Action Recognition

Temporal Transformer Networks with Self-Supervision for Action Recognition

Learning for Unconstrained Space-Time Video Super-Resolution

3DBodyNet: Fast Reconstruction of 3D Animatable Human Body Shape From a Single Commodity Depth Camera

Contact Info

Product

Resources

About