Google's TPU supercomputers train deep neural networks 50x faster than general-purpose supercomputers running a high-performance computing benchmark.

This paper deals with designing sensing matrix for compressive sensing systems. Traditionally, the optimal sensing matrix is designed so that the Gram of the equivalent dictionary is as close as possible to a target Gram with small mutual coherence. A novel design strategy is proposed, in which, unlike the traditional approaches, the measure considers of mutual coherence behavior of the equivalent dictionary as well as sparse representation errors of the signals. The optimal sensing matrix is defined as the one that minimizes this measure and hence is expected to be more robust against sparse representation errors. A closed-form solution is derived for the optimal sensing matrix with a given target Gram. An alternating minimization-based algorithm is also proposed for addressing the same problem with the target Gram searched within a set of relaxed equiangular tight frame Grams. The experiments are carried out and the results show that the sensing matrix obtained using the proposed approach outperforms those existing ones using a fixed dictionary in terms of signal reconstruction accuracy for synthetic data and peak signal-to-noise ratio for real images.

The three-dimensional (3D) MoS2 hierarchical nanospheres which assembled spontaneously by two-dimensional (2D) lamina have been successfully designed and fabricated in large scale via a simple hydrothermal process. Subsequently, the electromagnetic wave absorption properties of hierarchical MoS2 nanospheres compounded with polyvinylidene fluoride (PVDF) were investigated in a broad frequency range of 2–40 GHz. The results indicated that the MoS2/PVDF nanocomposites possess adjustable and enhanced wave absorption performance. Furthermore, the MA performance can be effectively tuned by the absorber’s thickness and filler content. In addition, the peculiar hierarchical nanostructure of MoS2 is beneficial to microwave absorption property compared with the bulk MoS2 and micrometer-sized MoS2. Moreover, the main microwave absorption mechanism including various polarization, destructive interference theory, and multiple reflection has been described in detail.

Recently, deep learning based video super-resolution (SR) methods have achieved promising performance. To simultaneously exploit the spatial and temporal information of videos, employing 3-dimensional (3D) convolutions is a natural approach. However, straight utilizing 3D convolutions may lead to an excessively high computational complexity which restricts the depth of video SR models and thus undermine the performance. In this paper, we present a novel fast spatio-temporal residual network (FSTRN) to adopt 3D convolutions for the video SR task in order to enhance the performance while maintaining a low computational load. Specifically, we propose a fast spatio-temporal residual block (FRB) that divide each 3D filter to the product of two 3D filters, which have considerably lower dimensions. Furthermore, we design a cross-space residual learning that directly links the low-resolution space and the high-resolution space, which can greatly relieve the computational burden on the feature fusion and up-scaling parts. Extensive evaluations and comparisons on benchmark datasets validate the strengths of the proposed approach and demonstrate that the proposed network significantly outperforms the current state-of-the-art methods.

We report laboratory experiments to investigate the dynamic failure characteristics of outburst‐prone coal using a split Hopkinson pressure bar (SHPB). For comparison, two groups of experiments are completed on contrasting coals—the first outburst‐prone and the second outburst‐resistant. The dynamic mechanical properties, failure processes, and energy dissipation of both outburst‐prone and outburst‐resistant coals are comparatively analyzed according to the obtained dynamic compressive and tensile stress‐strain curves. Results show that the dynamic stress‐strain response of both outburst‐prone and outburst‐resistant coal specimens comprises stages of compression, linear elastic deformation, then microfracture evolution, followed by unstable fracture propagation culminating in rapid unloading. The mechanical properties of both outburst‐prone and outburst‐resistant coal specimens exhibit similar features: The uniaxial compressive strength and indirect tensile strength increase linearly with the applied strain rate, and the peak strain increases nonlinearly with the strain rate, whereas the elastic modulus does not exhibit any clear strain rate dependency. Differences in the dynamic failure characteristics between outburst‐prone and outburst‐resistant coals also exist. The hardening effect of strain rate on outburst‐prone coal is more apparent than on outburst‐resistant coal, which is reflected in the dynamic increase factor at the same strain rate. However, the dynamic strength of outburst‐prone coals is still lower than that of outburst‐resistant coals due to its low quasi‐static strength. The dissipated energy of outburst‐prone coal is smaller than that of outburst‐resistant coal. Therefore, the outburst‐prone coal, characterized by low strength, high deformability, and small energy dissipation when dynamically loaded to failure, is more favorably disposed to the triggering and propagation of gas outbursts.