In skeleton-based human action recognition, spatial-temporal graph convolution networks (ST-GCNs) have achieved remarkable performances recently. However, how to explore more discriminative spatial and temporal features is still an open problem. The temporal graph convolution of the traditional ST-GCNs utilizes only one fixed kernel which cannot completely cover all the important stages of each action execution. Besides, the spatial and temporal graph convolution layers (GCLs) are serial connected, which mixes information of different domains and limits the feature extraction capability. In addition, the input features like joints, bones, and their motions are modeled in existing methods, but more input features are needed for better performance. To this end, this article proposes a novel multi-stream and enhanced spatial-temporal graph convolution network (MS-ESTGCN). For each basic block of MS-ESTGCN, densely connected multiple temporal GCLs with different kernel sizes are employed to aggregate more temporal features. To eliminate the adverse impact of information mixing, an additional spatial GCL branch is added to the block and the spatial features can be enhanced. Furthermore, we extend the input features by employing relative positions of joints and bones. Consequently, there are totally six data modalities (joints, bones, their motions and relative positions) that can be fed into the network independently with a six-stream paradigm. The proposed method is evaluated on two large scale datasets: NTU-RGB+D and Kinetics-Skeleton. The experimental results show that our method using only two data modalities delivers state-ofthe-art performance, and our methods using four and six data modalities further exceed other methods with a significant margin.INDEX TERMS Skeleton based action recognition, graph convolution network, multi-stream, enhanced spatial-temporal.
In the skeleton-based human action recognition domain, the spatial-temporal graph convolution networks (ST-GCNs) have made great progress recently. However, they use only one fixed temporal convolution kernel, which is not enough to extract the temporal cues comprehensively. Moreover, simply connecting the spatial graph convolution layer (GCL) and the temporal GCL in series is not the optimal solution. To this end, we propose a novel enhanced spatial and extended temporal graph convolutional network (EE-GCN) in this paper. Three convolution kernels with different sizes are chosen to extract the discriminative temporal features from shorter to longer terms. The corresponding GCLs are then concatenated by a powerful yet efficient one-shot aggregation (OSA) + effective squeeze-excitation (eSE) structure. The OSA module aggregates the features from each layer once to the output, and the eSE module explores the interdependency between the channels of the output. Besides, we propose a new connection paradigm to enhance the spatial features, which expand the serial connection to a combination of serial and parallel connections by adding a spatial GCL in parallel with the temporal GCLs. The proposed method is evaluated on three large scale datasets, and the experimental results show that the performance of our method exceeds previous state-of-the-art methods.
Roof-breaking characteristics and ground pressure behavior of the coalface are instrumental in guiding deep Jurassic coal seam mining, in particular in the Shaanxi and Inner Mongolia regions of China. A thick-plate mechanical model (TPMM) of the main roof was developed and applied to the case study of 21102 first-mined coalface (FMC) of the Hulusu Coal Mine (HCM) in the Hujirt Mining Area (HMA), China. A theoretical analysis performed via the developed model revealed that the first and periodic breaking intervals of the main roof were 40.6 and 25.0 m, respectively. The roof failure occurred in the tensile mode, was controlled by the internal stress
σ
x
in the rock strata, and started from the center of the long side with the fixed support in the goaf. The field measurement of roof weighting was also performed for the coalface advance from zero to 400 m. The measurement results showed that the first weighting average interval was 41.4 m, and the average interval of periodic weighting was 22.0 m, which agreed with the theoretical calculation and proved the proposed model’s feasibility. Finally, the frequency distribution features of the hydraulic support working resistance in the FMC were analyzed statistically. The results showed that the ZY10000-16/32D supports could adapt to the mining geological conditions of the FMC. However, the margin of the rated working resistance of supports was still small. Thus, roof management enhancement during the mining process was strongly recommended. These research findings could offer theoretical guidance for safe and high-efficiency production in the coal mines under similar geological conditions.
Summary
Squirt flow plays an essential role in elastic modulus dispersion and attenuation for fluid-saturated cracked porous rocks. Mavko-Jizba model and relevant modified models can describe the squirt flow well based on the related elastic moduli, such as dry/drained bulk modulus. However, when these elastic moduli are challenging to attain, it is impossible to model the squirt-flow-related elastic moduli and attenuations with the models. On the other hand, the effective medium theory (EMT) model can estimate these elastic moduli, but cannot predict the undrained/relaxed and partially-relaxed saturated elastic moduli and the squirt-flow-related attenuations. This paper extended an EMT model—Cracks-Pores Effective Medium (CPEM) model to cover the undrained/relaxed and partially-relaxed states following the elastic-viscoelastic correspondence principle. The proposed model (i.e., the CPEMF model) can thus estimate the elastic moduli over the different states (dry/drained, undrained/relaxed, partially-relaxed, unrelaxed) and associated attenuations. It agrees well with the prediction of the modified Mavko-Jizba-Gurevich model (MJGZ-HF) at unrelaxed state and is precisely consistent with the prediction of Gassmann at undrained/relaxed state. Also, it analytically shows good consistency with the modified Mavko-Jizba-Gurevich model (MJGZ-MF) at partially-relaxed state. The numerical simulations of CPEM/CPEMF models and MJGZ-HF/ MJGZ-MF models show good agreement at the different states. Furthermore, we interpreted the experimental data on a basaltic sample and a sandstone sample with the CPEM/CPEMF models. The CPEMF model's predictions of elastic modulus at different states and associated modulus dispersion/attenuation are in good agreement with the corresponding measured ones, suggesting that the proposed CPEMF model can efficiently predict the elastic moduli at different states (dry/drained, undrained/relaxed, partially relaxed, unrelaxed) and quantify the squirt-flow-related elastic modulus dispersion and attenuation among different states well.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.