Search

Skeleton-based human action recognition has made great progress, especially with the development of a graph convolution network (GCN). The most important work is ST-GCN, which automatically learns both spatial and temporal patterns from skeleton sequences. However, this method still has some imperfections: only short-range correlations are appreciated, due to the limited receptive field of graph convolution. However, long-range dependence is essential for recognizing human action. In this work, we propose the use of a spatial-temporal relative transformer (ST-RT) to overcome these defects. Through introducing relay nodes, ST-RT avoids the transformer architecture, breaking the inherent skeleton topology in spatial and the order of skeleton sequence in temporal dimensions. Furthermore, we mine the dynamic information contained in motion at different scales. Finally, four ST-RTs, which extract spatial-temporal features from four kinds of skeleton sequence, are fused to form the final model, multi-stream spatial-temporal relative transformer (MSST-RT), to enhance performance. Extensive experiments evaluate the proposed methods on three benchmarks for skeleton-based action recognition: NTU RGB+D, NTU RGB+D 120 and UAV-Human. The results demonstrate that MSST-RT is on par with SOTA in terms of performance.

This paper proposes an algorithm for transmitting and reconstructing the estimated point cloud by temporally estimating a dynamic point cloud sequence. When a non-rigid 3D point cloud sequence (PCS) is input, the sequence is divided into groups of point cloud frames (PCFs), and a key PCF is selected. The 3D skeleton is predicted through 3D pose estimation, and the motion of the skeleton is estimated by analyzing the joints and bones of the 3D skeleton. For the deformation of the non-rigid human PC, the 3D PC model is transformed into a mesh model, and the key PCF is rigged using the 3D skeleton. After deforming the key PCF into the target PCF utilizing the motion vector of the estimated skeleton, the residual PC between the motion compensation PCF and the target PCF is generated. If there is a key PCF, the motion vector of the target PCF, and a residual PC, the target PCF can be reconstructed. Just as compression is performed using pixel correlation between frames in a 2D video, this paper compresses 3D PCFs by estimating the non-rigid 3D motion of a 3D object in a 3D PC. The proposed algorithm can be regarded as an extension of the 2D motion estimation of a rigid local region in a 2D plane to the 3D motion estimation of a non-rigid object (human) in 3D space. Experimental results show that the proposed method can successfully compress 3D PC sequences. If it is used together with a PC compression technique such as MPEG PCC (point cloud compression) in the future, a system with high compression efficiency may be configured. [ABSTRACT FROM AUTHOR]

Copyright of Journal of South China Normal University (Natural Science Edition) / Huanan Shifan Daxue Xuebao (Ziran Kexue Ban) is the property of Journal of South China Normal University (Natural Science Edition) Editorial Office and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Existing 3D skeleton-based action recognition approaches reach impressive performance by encoding handcrafted action features to image format and decoding by CNNs. However, such methods are limited in two ways: a) the handcrafted action features are difficult to handle challenging actions, and b) they generally require complex CNN models to improve action recognition accuracy, which usually occur heavy computational burden. To overcome these limitations, we introduce a novel AFE-CNN , which devotes to enhance the features of 3D skeleton-based actions to adapt to challenging actions. We propose feature enhance modules from key joint, bone vector, key frame and temporal perspectives, thus the AFE-CNN is more robust to camera views and body sizes variation, and significantly improve the recognition accuracy on challenging actions. Moreover, our AFE-CNN adopts a light-weight CNN model to decode images with action feature enhanced, which ensures a much lower computational burden than the state-of-the-art methods. We evaluate the AFE-CNN on three benchmark skeleton-based action datasets: NTU RGB + D, NTU RGB + D 120, and UTKinect-Action3D, with extensive experimental results demonstrate our outstanding performance of AFE-CNN. [ABSTRACT FROM AUTHOR]

The 3D structured Zn anode which is grown on the surface of carbon cloth/carbon nanofiber framework (CC-CNF) can reduce the nucleation overpotential of the electrode and suppress dendrites owing to the high conductivity and abundant nucleation sites of the 3D skeleton. [Display omitted] Zn-ion battery (ZIB) is a promising electrochemical energy storage device because of the high performance, safety, and economic benefits, etc. However, the formation of Zn dendrites at anodes is seriously depressed their cycling life, Coulombic efficiency, safety, and capacity. Inhibition of Zn dendrites is considered as an efficient solution to improving the performance of ZIB. Herein, we demonstrate a 3D structured Zn anode coated in carbon nanofiber framework with carbon cloth as substrate (Zn@CC-CNF), which has the abundant Zn nucleation sites and homogeneous electrical field distribution for uniform Zn deposition, low nucleation/deposition energy barrier, and the capacity of alleviating side reactions. Thus, the Zn@CC-CNF based symmetric cells is stability operated over 400 h, which is over 400% longer than that of bare Zn coated on carbon cloth (Zn@CC). The impressive performance of the Zn@CC-CNF anode also endows the assembled Zn-MnO 2 full cell with long-term cycling stability. Better yet, the intrinsic flexibility of the carbon substrate enables the fabrication of flexible Zn-MnO 2 cell with gratifying mechanical deformation tolerance, illustrating their great application potential in flexible energy storage. [ABSTRACT FROM AUTHOR]

In the field of body affect recognition, the majority of literature is based on experiments performed on datasets where trained actors simulate emotional reactions. These acted and unnatural expressions differ from the more challenging genuine emotions, thus leading to less valuable results. In this article, a solution for basic non-acted emotion recognition based on 3D skeleton and Deep Neural Networks (DNNs) is provided. The proposed work introduces three majors contributions. First, temporal local movements performed by subjects are examined frame-by-frame, unlike the current state-of-the-art in non-acted body affect recognition where only static or global body features are considered. Second, an original set of global and time-dependent features for body movement description is provided. Third, this is one of the first works to use deep learning methods in the current non-acted body affect recognition literature. Due to the novelty of the topic, only the UCLIC dataset is currently considered the benchmark for comparative tests. On the latter, the proposed method outperforms all the competitors. [ABSTRACT FROM AUTHOR]

Although solid composite electrolytes show tremendous potential for the practical solid‐state lithium metal batteries, searching for a straightforward tactic to promote the ion conduction at electrolyte/electrode interface, especially settling lithium dendrites formation caused by the concentration gradient polarization, are still long‐standing problems. Here, the authors report a corrugated 3D nanowires‐bulk ceramic‐nanowires (NCN) skeleton reinforced composite electrolyte with regulated interfacial Li‐ion transport behavior. The special and integrated NCN skeleton endows the electrolyte with fast Li‐ion transfer and solves the Li+ concentration polarization at electrode/electrolyte interface, thereby eliminating the energy barrier originated from the redistribution of charge carriers and offering homogeneous interfacial Li‐ion flux on lithium anode. As a "double insurance", the bulk ceramic sheet in 3D framework enables the electrolyte to block the mobility of anions. The rational designed NCN composite electrolyte exhibits excellent ionic conductivity and the assembled all‐solid‐state battery possesses 90.2% capacity retention after 500 cycles. The proposed strategy affords a special insight in designing high‐performance solid composite electrolytes. [ABSTRACT FROM AUTHOR]

Recent advances in terahertz (THz) absorbing materials and technology show futuristic potentials for practical applications in THz radars and telecommunications, stealth and shielding. However, the lack of versatile materials naturally working in this specific electromagnetic wave region with simultaneously featuring high absorption efficiencies, ultrabroad bandwidths, low‐costs, good stabilities, and flexibilities, is impeding the proliferation of real THz disruptive applications. Here a kind of flexible structure material, 3D nickel (Ni) skeleton, fabricated from an electroplating sintering method with irregular pore distribution makes possible the successful realization of a highly absorbing response for ultrabroadband THz waves due to the effective combination of both material and structural absorption mechanisms. 3D Ni skeletons with 90 ppi nonuniform pore‐size ranges enable >99% absorption capabilities in the frequency range of 0.5–2.0 THz independent on both the THz incidence angles and polarizations. Experimental validation of THz shielding implemented on both 100 GHz and 4.3 THz video imaging systems corroborates the highly efficient absorbing with frequency expansibility. Such capabilities are further verified on millimeter‐wave security checkers for 32–36 GHz. This prototypical demonstration lays the foundation for the next‐generation THz absorbing technology, accelerating advanced THz technologies toward practical applications. [ABSTRACT FROM AUTHOR]

This paper addresses the challenge of 3D motion recovery by exploiting the spatio-temporal correlations of corrupted 3D skeleton sequences. We propose a new 3D motion recovery method using spatio-temporal reconstruction, which uses joint low-rank and sparse priors to exploit temporal correlation and an isometric constraint for spatial correlation. The proposed model is formulated as a constrained optimization problem, which is efficiently solved by the augmented Lagrangian method with a Gauss–Newton solver for the subproblem of isometric optimization. The experimental results on the CMU motion capture dataset, Edinburgh dataset, and two Kinect datasets demonstrate that the proposed approach achieves better motion recovery than the state-of-the-art methods. The proposed method is applicable to Kinect-like skeleton tracking devices and pose estimation methods that cannot provide accurate estimation of complex motions, especially in the presence of occlusion. [ABSTRACT FROM AUTHOR]

This paper proposes a dual-stream 3D space-time convolutional neural network action recognition framework. The original depth map sequence data is set as the input in order to study the global space-time characteristics of each action category. The high correlation within the human action itself is considered in the time domain, and then the deep motion map sequence is introduced as the input to another stream of the 3D space-time convolutional network. Furthermore, the corresponding 3D skeleton sequence data is set as the third input of the whole recognition framework. Although the skeleton sequence data has the advantage of including 3D information, it is also confronted with the problems of the existence of rate change, temporal mismatch and noise. Thus, specially designed space-time features are applied to cope with these problems. The proposed methods allow the whole recognition system to fully exploit and utilize the discriminatory space-time features from different perspectives, and ultimately improve the classification accuracy of the system. Experimental results on different public 3D data sets illustrate the effectiveness of the proposed method. [ABSTRACT FROM AUTHOR]

Using 3D models of bones can highly improve accuracy and reliability of orthopaedic evaluation. However, it may impose excessive computational load. This article proposes a fully automatic method for extracting a compact model of the femur from its 3D model. The proposed method works by extracting a 3D skeleton based on the clinical parameters of the femur. Therefore, in addition to summarizing a 3D model of the bone, the extracted skeleton would preserve important clinical and anatomical information. The proposed method has been applied on 3D models of 10 femurs and the results have been evaluated for different resolutions of data. [ABSTRACT FROM AUTHOR]

Uneven deposition and serious volume change restrict the commercial application of Li metal anode. Sn based alloy is pursued as the advanced anode via an alloying/dealloying mechanism while considering its resource abundance, low price, and high specific capacity. Herein, a three-dimensional (3D) Li 22 Sn 5 alloy skeleton is integrated with Li metal phase via phase-segregation process by a facial melting method, in which Li-rich dual-phase Li-Sn alloy named SnLi90 is produced. The free Li phase accounts for a huge proportion of the electrode weight and acts as the active material offering ultra-high specific capacity of 3087 mAh g−1. Alternatively, the unique interconnected rod-like Li 22 Sn 5 network serves as electrochemical "inert" framework due to its much higher redox potential. The favorable Li affinity of the Li 22 Sn 5 skeleton contributes to conformal Li deposition. Consequently, the SnLi90 anode can cycle stably for 1500 h in ester electrolyte at 1 mA cm−2 and 1 mAh cm−2, almost three times longer than that of the bare Li anode. The SnLi90 anode contains two phases, i.e., Li metal and Li 22 Sn 5 alloy. Li enables high specific capacity. Li 22 Sn 5 skeleton maintains structural integrity, electrochemical "inertness" and induces conformal deposition of Li. [Display omitted] [ABSTRACT FROM AUTHOR]

Skeleton-based Human Activity Recognition has achieved great interest in recent years as skeleton data has demonstrated being robust to illumination changes, body scales, dynamic camera views, and complex background. In particular, Spatial–Temporal Graph Convolutional Networks (ST-GCN) demonstrated to be effective in learning both spatial and temporal dependencies on non-Euclidean data such as skeleton graphs. Nevertheless, an effective encoding of the latent information underlying the 3D skeleton is still an open problem, especially when it comes to extracting effective information from joint motion patterns and their correlations. In this work, we propose a novel Spatial–Temporal Transformer network (ST-TR) which models dependencies between joints using the Transformer self-attention operator. In our ST-TR model, a Spatial Self-Attention module (SSA) is used to understand intra-frame interactions between different body parts, and a Temporal Self-Attention module (TSA) to model inter-frame correlations. The two are combined in a two-stream network, whose performance is evaluated on three large-scale datasets, NTU-RGB+D 60, NTU-RGB+D 120, and Kinetics Skeleton 400, consistently improving backbone results. Compared with methods that use the same input data, the proposed ST-TR achieves state-of-the-art performance on all datasets when using joints' coordinates as input, and results on-par with state-of-the-art when adding bones information. • A model for action recognition employing spatial–temporal Transformer attention. • SSA substitutes graph convolution exploiting dynamic inter-frame relation discovering. • TSA substitutes temporal convolution introducing long-range dynamic reasoning. • The model is lightweight and adapts to different networks and input configurations. • We outperform ST-GCN and A-GCN on joint based prediction on NTU and Kinetics-400. [ABSTRACT FROM AUTHOR]