Search

ZnSeTe quantum dots (QDs) offer an efficient avenue for realizing heavy-metal-free light-emitting diodes (LEDs) that meet the Rec.2100 blue standard. Synthetic core–shell engineering has enabled big advances in the external quantum efficiency (EQE) of ZnSeTe QD-LEDs. However, the mechanisms behind the degradation of the operational stability of ZnSeTe QD-LEDs remain relatively unexplored. In this study, we explore the impact of ligand density and composition on both material and device stability. We developed a solid-film ligand exchange utilizing an inorganic X-type ligand (zinc chloride), revealing that the substitution of inorganic ligands for organic counterparts significantly influences the stability of both materials and devices.

Video analytics at mobile edge servers offers significant benefits like reduced response time and enhanced privacy. However, guaranteeing various quality-of-service (QoS) requirements of dynamic video analysis requests on heterogeneous edge devices remains challenging. In this paper, we propose a scalable online video analytics scheme, called Novas, which automatically makes precise service configuration adjustments upon constant video content changes. Specifically, Novas leverages the filtered confidence sum and a two-window t-test to online detect accuracy fluctuations without ground truth information. In such cases, Novas efficiently estimates the performance of all potential service configurations through a singular value decomposition (SVD)-based collaborative filtering method. Finally, given the NP-hardness of the optimal scheduling problem, a heuristic scheduling strategy that maximizes the minimum remaining resources is devised to schedule the most suitable configurations to servers for execution. We evaluate the effectiveness of Novas through extensive hybrid experiments conducted on a dedicated testbed. Results show that Novas can achieve a substantial over 27$\times$× improvement in satisfying the accuracy requirements compared with existing methods adopting fixed configurations, while ensuring latency requirements. Moreover, Novas improves the goodput of the system by an average of 37.86% compared to existing state-of-the-art scheduling solutions.

Batch normalization (BN) enhances the training of deep ReLU neural network with a composition of mean centering (centralization) and variance scaling (unitization). Despite the success of BN, there lacks a theoretical explanation to elaborate the effects of BN on training dynamics and guide the design of normalization methods. In this paper, we elucidate the effects of centralization and unitization on training deep ReLU neural networks for BN. We first reveal that feature centralization in BN stabilizes the correlation coefficients of features in unnormalized ReLU neural networks to achieve feature decorrelation and accelerate convergence in training. We demonstrate that weight centralization that subtracts means from weight parameters is equivalent to BN in feature decorrelation and achieves the same linear convergence rate in training. Subsequently, we show that feature unitization in BN enables dynamic learning rate that inversely varies with the norm of features for training and propose an adaptive loss function to emulate feature unitization. Furthermore, we exemplify the theoretical results to develop an efficient alternative to BN using a simple combination of weight centralization and the proposed adaptive loss function. Extensive experiments show that the proposed method achieves comparable classification accuracy and evidently reduces memory consumption in comparison to BN, and outperforms normalization-free methods in image classification. We further extend the weight centralization to enable small-batch training for object detection networks.

The 3-D boundary identification of magnetic targets is of significant importance for the magnetic characteristics assessment of magnetic devices. A high-fidelity reconstruction method for the 3-D boundary of tiny-sized magnetic targets through inversion techniques is presented in this article. Based on nitrogen-vacancy (NV) color centers in diamond, vector magnetic imaging with high sensitivity, wide field of view, and alignment-free is realized. Based on the anomaly distribution of the total magnetic field intensity (TMI), we determined the spatial dimensions and depth for the inversion and discretized the 3-D space into a grid. In order to overcome magnetic field noise and overlapping interference, we proposed a multiwindow inversion model by constructing inversion windows of various sizes. The application of various window sizes for inversion, along with the utilization of their corresponding gradient weights in a weighted combination, effectively addresses the ill-conditioned inversion challenges stemming from the nonlinear crosstalk of magnetic field noise. The experimental results show that the proposed method can achieve an inversion accuracy of 99.1% with a sensitivity of $2~\mu $ T/ $\surd $ Hz. Compared with the method of L0–L2 norm regularization inversion, the inversion accuracy is improved by 3.5%. This method opens up a way for high-fidelity 3-D magnetic target boundary inversion.

Recent learned image compression models surpass manually designed methods in rate-distortion performance by introducing nonlinear transforms and end-to-end optimization. However, there still lack quantitative measurements that efficiently evaluate the latent representations inferred by learned image compression models. To address this problem, we develop novel measurements on robustness and importance of the latent representations. We first propose an admissible range that can be efficiently estimated via gradient ascent and descent for establishing the empirical distribution of latent representations. Consequently, the in-distribution region within the admissible range is derived to measure the robustness and channel importance of latent representations of natural images. Visualization demonstrates the statistics of latent representations are significantly distinguishing in the properties of robustness and linearity within and outside the in-distribution region. To our best knowledge, this paper proposes the first statistically meaningful measurements for learned image compression and successfully applies the measurements in corruption alleviation during successive image compression and post-training pruning in a training-free fashion. Compared with existing methods, the shrunk in-distribution constraint derived from the in-distribution region achieves superior robustness and rate-distortion performance in successive compression. The channel importance allows post-training pruning to achieve comparable rate-distortion performance with a reduction of up to 60% entropy coding time.

Scattering networks on Euclidean domains are capable of analytically realizing signal representation invariant to transformations such as translation, rotation and scaling with wavelets. However, existing scattering networks defined on the sphere and Riemannian manifolds only consider axisymmetric wavelets and are restricted in representation by the isotropic filter structures. In this paper, we propose a novel anisotropic spherical scattering network to achieve multi-scale directional representation for spherical signals. The scattering transform is realized by cascading directional spin wavelets and modulus operators to propagate localized signal components of varying scales and directions in a recursive manner without parameter tuning. Furthermore, a combined spherical scattering network is presented to guarantee the invariance to arbitrary rotations about the z-axis by incorporating the scattering coefficients along the dimension of rotation angle. To our best knowledge, the proposed network is the first to achieve multi-scale anisotropic filtering via the scattering transform on the sphere. We demonstrate in theory that the proposed network is energy preserving, invariant to azimuthal rotations and stable to diffeomorphisms. Extensive experimental results on benchmark datasets show that the proposed network achieves state-of-the-art performance in spherical signal analysis on various 2-D and 3-D datasets mapped to the sphere.

Graphical abstract:

Model quantization is essential to deploy deep convolutional neural networks (DCNNs) on resource-constrained devices. In this article, we propose a general bitwidth assignment algorithm based on theoretical analysis for efficient layerwise weight and activation quantization of DCNNs. The proposed algorithm develops a prediction model to explicitly estimate the loss of classification accuracy led by weight quantization with a geometrical approach. Consequently, dynamic programming is adopted to achieve optimal bitwidth assignment on weights based on the estimated error. Furthermore, we optimize bitwidth assignment for activations by considering the signal-to-quantization-noise ratio (SQNR) between weight and activation quantization. The proposed algorithm is general to reveal the tradeoff between classification accuracy and model size for various network architectures. Extensive experiments demonstrate the efficacy of the proposed bitwidth assignment algorithm and the error rate prediction model. Furthermore, the proposed algorithm is shown to be well extended to object detection. [ABSTRACT FROM AUTHOR]

Graphical abstract:

Graphical abstract:

C7H4CuF6NO4, triclinic, P1̄ (no. 2), a= 8.5122(5) Å, b= 8.5458(6) Å, c= 8.8674(6) Å, α= 75.042(3)°, β= 68.601(2)°, γ= 89.809(3)°, V= 577.26(7) Å3, Z= 2, Rgt(F) = 0.0247, wRref(F2) = 0.0625, T= 152.41 K.

The previously derived marine turbulence modulation transfer function (MTF) adopts the classical assumption of isotropic and Kolmogorov turbulence. However, more theoretical research shows that marine turbulence demonstrates anisotropic non-Kolmogorov properties. In this study, new long-exposure plane and spherical wave MTF models under anisotropic non-Kolmogorov marine turbulence will be investigated. Comparative analyses are performed for the new models and the existing ones that have been developed under the isotropic marine turbulence and the anisotropic terrestrial turbulence cases. It turns out that the anisotropic marine turbulence influences the electro-optical imaging system less than the isotropic marine turbulence does. Also, under the same turbulence strength condition, due to the humidity fluctuations of marine atmosphere, the anisotropic marine turbulence will reduce more severely the quality of electro-optical imaging systems than the anisotropic terrestrial turbulence. Investigations conducted in this paper are beneficial to understand better the electro-optical imaging mechanism in marine atmospheric turbulence media.

Laparoscopic liver resection under hemihepatic vascular inflow control has advantages over Pringle’s maneuver, especially in patients with cirrhosis. From January 2016 to August 2016, 7 patients who underwent total laparoscopic left hepatectomy under hemihepatic vascular inflow occlusion using the extra-glissonian approach were included in this study. All were hepatitis B carriers and 4 had cirrhosis. The mean operation time was 247 minutes. The mean transection time was 110 minutes. No patient needed additional Pringle’s maneuver. The mean intraoperative blood loss was 74 ml and no patient required blood transfusion. No open conversion happened. Postoperatively, no patient developed complications and there was no perioperative mortality. The mean resection margin was 2 cm. The mean hospital stay was 6 days. Upon a mean follow-up of 9 months, no patient developed tumor recurrence. The technique of total laparoscopic left hepatectomy using extra-glissonian approach was safe and feasible. The early surgical outcomes were good.

At present, microneedles have been widely used in transdermal drug delivery and detection due to their painless and non-invasive advantages, among which porous microneedles have great potential in dry or liquid drug delivery and disease monitoring due to their interoperable pore structure. However, previous literature paid little attention to the porous microneedles, and worse, most of them focused on the preparation methods of the porous microneedles used for drug delivery, while the matrix materials of porous microneedles and the potential for biomarker detection were rarely mentioned. The purpose of this review is to summarize the commonly used matrix materials for porous microneedles and highlight the ongoing challenges and prospects of their application in transdermal drug delivery and disease detection.

With advances in technology, laparoscopic liver resection is widely accepted. Laparoscopic liver resection under hemihepatic vascular inflow occlusion has advantages over the conventional total hepatic inflow occlusion using the Pringle's maneuver, especially in patients with cirrhosis.

Brugia malayi is one of the parasitic worms which causes lymphatic filariasis in humans. Its geographical distribution includes a large part of Asia. Despite its wide distribution, very little is known about the genetic variation and molecular epidemiology of this species. In this study, the internal transcribed spacer 1 (ITS1) nucleotide sequences of B. malayi from microfilaria-positive human blood samples in Northeast Borneo Island were determined, and compared with published ITS1 sequences of B. malayi isolated from cats and humans in Thailand. Multiple alignment analysis revealed that B. malayi ITS1 sequences from Northeast Borneo were more similar to each other than to those from Thailand. Phylogenetic trees inferred using Neighbour-Joining and Maximum Parsimony methods showed similar topology, with 2 distinct B. malayi clusters. The first cluster consisted of Northeast Borneo B. malayi isolates, whereas the second consisted of the Thailand isolates. The findings of this study suggest that B. malayi in Borneo Island has diverged significantly from those of mainland Asia, and this has implications for the diagnosis of B. malayi infection across the region using ITS1-based molecular techniques. [ABSTRACT FROM PUBLISHER]

In order to overcome the defects of the traditional urban park green space system planning methods, accessibility method is introduced to the planning of urban park green space system in this paper. Based on proposing the necessity of urban green space accessibility evaluation, this paper introduces principles of accessibility evaluation and constructs the evaluation model based on the cost resistance mode. The paper evaluates the accessibility service scope of urban park green space in Luancheng city on the basis of the data collection and field surveys supported by ArcGIS technology, then based on this，it puts forward urban park green space system planning layout scheme and optimization measures.The study has shown that accessibility method makes up for the deficiencies of the traditional urban park green space system planning methods and has a practical guidance.

Large collapse of underground house of a hydroelectric power station happened while excavation of the top arch. The volume of collapse was up to 5000m3. With consideration of the field geological conditions and practical construction, by strength reduction finite element method (FEM) and discrete element method (DEM), the process of collapse is simulated, which is compared with the field monitoring data. The mechanism of instability failure is analyzed. Some conclusions are obtained. Firstly, the development of fractured weak zone β80, faults, joints and fissures above the top arch are the basic factors to induce this collapse. Secondly, abrupt blasting without supporting is the direct factor. So collapse of underground house can be avoided by reinforcement and supporting on the geological weak zone in time.

For the cast-steel joint of large span roof system, the main difficulties of prototype stress test encountered are the large scale of the structure system, large loading system and large supporting system, etc. The hydraulic jack of high tonnage is adopted to apply the load. The frame structure of channel steel is used to fix the cast-steel joint and the loading system. The reverse loading system is elaborated to simulate the axial tensile force. So prototype stress test on the cast-steel joint of large span roof system is accomplished successfully. The test results show that, under the designing load condition, cast-steel joint is in a linear elastic state, the structure is globally safe; the maximum surface stress of cast-steel joint is adjacent to the crossing point of steel pipes; the stress value of test is well coincident with one of finite element analysis. This test device system is especially suitable for the stress state test of large components, and can also provide important reference to the similar projects.

AbstractWe examine the solution procedure for the inventory model with partially backlogging. Through the rectangular coordinate system, we construct a family of horizon line and discover the properties for these horizon lines. We find the criterion to distinguish the minimum solution into interior minimum problem and boundary minimum problem. Our algorithm is self-interesting and motivated to the practitioners in the operational research.

When a natural porous formation is viewed from an Eulerian perspective, incomplete characterization of the hydraulic conductivity leads to nonlocality in the constitutive theory, irrespective of whether the medium has evolving heterogeneity with velocity fluctuations over all scales. Within this framework two nonlocal constitutive models are developed for natural systems, one for conservative tracers and the other for reactive chemicals undergoing linear nonequilibrium sorption with deterministic rate constants. Exact solutions for the mean concentrations are obtained and used to determine the mean values of the spatial moments up to the third. A Lagrangian model (Dagan and Cvetkovic, 1993) for a similar problem is reviewed and comparisons are made between the expected Lagrangian and Eulerian moments. If the local-scale dispersive process is neglected in the Eulerian analysis, then the Lagrangian moments obtain. However, if the local-scale dispersive process is included in the Eulerian model, then the second transverse moment disagrees with that obtained through the Lagrangian analysis. It is shown that this disagreement is especially acute in the asymptotic limits.

Asymptomatic severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) infections are well documented. Healthcare workers (HCW) are at increased risk of infection due to occupational exposure to infected patients. We aim to determine the prevalence of SARS-CoV-2 antibodies among HCW who did not come to medical attention.