Search

Spatially localized deformation components are very useful for shape analysis and synthesis in 3D geometry processing. Several methods have recently been developed, with an aim to extract intuitive and interpretable deformation components. However, these techniques suffer from fundamental limitations especially for meshes with noise or large-scale nonlinear deformations, and may not always be able to identify important deformation components. In this paper we propose a mesh-based variational autoencoder architecture that is able to cope with meshes with irregular connectivity and nonlinear deformations, assuming that the analyzed dataset contains meshes with the same vertex connectivity, which is common for deformation analysis. To help localize deformations, we introduce sparse regularization in this framework, along with spectral graph convolutional operations. Through modifying the regularization formulation and allowing dynamic change of sparsity ranges, we improve the visual quality and reconstruction ability of the extracted deformation components. Our system also provides a nonlinear approach to reconstruction of meshes using the extracted basis, which is more effective than the current linear combination approach. As an important application of localized deformation components and a novel approach on its own, we further develop a neural shape editing method, achieving shape editing and deformation component extraction in a unified framework, and ensuring plausibility of the edited shapes. Extensive experiments show that our method outperforms state-of-the-art methods in both qualitative and quantitative evaluations. We also demonstrate the effectiveness of our method for neural shape editing.

Input parallel output parallel (IPOP) DC/DC system is widely utilized in many occasions like data center and hybrid electric vehicle (HEV). However, the efficiency characteristics of different converters usually differs from each other, and the conventional control strategy has difficult to realize efficient conversion of overall power of the system. To address this issue, Buck converter is taken as the module and whose efficiency model under current continuous conduction mode (CCM) is presented. Meantime, parameter identification method based on least square (LS) algorithm is proposed as well. In order to realize the efficiency optimization of the whole parallel system, the constraint conditions of each module are analyzed, and the current distribution calculation method based on sequential quadratic programming (SQP) algorithm for the parallel system is introduced. On this basis, a control strategy of proportional current distribution (PCD) is proposed, which not only ensures that the parallel system can always run as efficiently as possible under different loads, but also realizes the flexible switching between the current-sharing mode and current distribution mode. Finally, the simulation and experimental results prove that the proposed method can achieve 4.46% higher efficiency than conventional method within the given power range of each converter, and realize the efficiency optimization of the IPOP DC/DC system to the greatest extent.

An improved modulation strategy based on minimum energy storage for DC-link capacitance reduction in a six-switch AC-AC converter is proposed. The proposed modulation strategy enables the energy on the capacitor to accumulate and release twice each in a complete switching cycle, achieving the effect of “fast charging and discharging”. Meanwhile, the inversion and rectification are modulated synchronously. Hence, there is minimum energy stored in the DC-link capacitor. Then, the time average modeling analysis is presented to take insight analysis. When there is the same voltage ripples constraint on the DC side for the conventional and improved modulation strategies, the six-switch converter under the improved modulation strategy has the much less capacitance value of the storage capacitor and even realizes non-electrolytic capacitance. Therefore, improving the system efficiency, power density, and output waveform quality and extending the system life can be achieved. The operation principle and modulation strategy are discussed in detail. Finally, the simulation model and experimental prototype are built to verify effectiveness of the topology and correctness of the proposed six-switch AC-AC converter modulation strategy.

An interleaved high-step-up current sharing DC–DC converter with coupled inductors is proposed in this paper. The operation principle and property of this converter are analyzed. The ripple of the input current in the proposed converter is decreased significantly by using the two-phase parallel interleaved input. The voltage gain is extended and the switch voltage stress is reduced by the coupled inductors. The leakage inductance of the coupled inductors reduces the reverse-recovery problem of the output diode, resulting in the reduction of reverse-recovery losses. As there are two interleaved phases in the proposed converter, the third winding of each coupled inductor is embedded in another phase. With this design, when the leakage inductance or duty cycle is asymmetrical, the current sharing performance is still positive. Consequently, the new topology is very suitable for applications to occasions with low voltage input and high voltage output, such as the fuel cell power system. Finally, the performance of this topological circuit is verified by a prototype with 500 W output.

Deformation component analysis is a fundamental problem in geometry processing and shape understanding. Existing approaches mainly extract deformation components in local regions at a similar scale while deformations of real-world objects are usually distributed in a multi-scale manner. In this paper, we propose a novel method to exact multiscale deformation components automatically with a stacked attention-based autoencoder. The attention mechanism is designed to learn to softly weight multi-scale deformation components in active deformation regions, and the stacked attention-based autoencoder is learned to represent the deformation components at different scales. Quantitative and qualitative evaluations show that our method outperforms state-of-the-art methods. Furthermore, with the multiscale deformation components extracted by our method, the user can edit shapes in a coarse-to-fine fashion which facilitates effective modeling of new shapes., 15 pages

Due to the abuse of antibiotics, antibiotic residues can be detected in both natural environment and various industrial products, posing threat to the environment and human health. Here we describe the design and implementation of an engineered Escherichia coli capable of degrading tetracycline (Tc)-one of the commonly used antibiotics once on humans and now on poultry, cattle and fisheries. A Tc-degrading enzyme, TetX, from the obligate anaerobe Bacteroides fragilis was cloned and recombinantly expressed in E. coli and fully characterized, including its Km and kcat value. We quantitatively evaluated its activity both in vitro and in vivo by UV–Vis spectrometer and LC-MS. Moreover, we used a tetracycline inducible amplification circuit including T7 RNA polymerase and its specific promoter PT7 to enhance the expression level of TetX, and studied the dose-response of TetX under different inducer concentrations. Since the deployment of genetically modified organisms (GMOs) outside laboratory brings about safety concerns, it is necessary to explore the possibility of integrating a kill-switch. Toxin-Antitoxin (TA) systems were used to construct a mutually dependent host-plasmid platform and biocontainment systems in various academic and industrious situations. We selected nine TA systems from various bacteria strains and measured the toxicity of toxins (T) and the detoxifying activity of cognate antitoxins (A) to validate their potential to be used to build a kill-switch. These results prove the possibility of using engineered microorganisms to tackle antibiotic residues in environment efficiently and safely. Keywords: TetX, Antibiotic residue, Toxin-antitoxin system, Kill-switch

We present a novel algorithm (DeepMNavigate) for global multi-agent navigation in dense scenarios using deep reinforcement learning (DRL). Our approach uses local and global information for each robot from motion information maps. We use a three-layer CNN that takes these maps as input to generate a suitable action to drive each robot to its goal position. Our approach is general, learns an optimal policy using a multi-scenario, multi-state training algorithm, and can directly handle raw sensor measurements for local observations. We demonstrate the performance on dense, complex benchmarks with narrow passages and environments with tens of agents. We highlight the algorithm’s benefits over prior learning methods and geometric decentralized algorithms in complex scenarios.

Due to the abuse of antibiotics, antibiotic residues can be detected in both natural environment and various industrial products, posing threat to the environment and human health. Here we describe the design and implementation of an engineered Escherichia coli capable of degrading tetracycline (Tc)-one of the commonly used antibiotics once on humans and now on poultry, cattle and fisheries. A Tc-degrading enzyme, TetX, from the obligate anaerobe Bacteroides fragilis was cloned and recombinantly expressed in E. coli and fully characterized, including its Km and kcat value. We quantitatively evaluated its activity both in vitro and in vivo by UV–Vis spectrometer and LC-MS. Moreover, we used a tetracycline inducible amplification circuit including T7 RNA polymerase and its specific promoter PT7 to enhance the expression level of TetX, and studied the dose-response of TetX under different inducer concentrations. Since the deployment of genetically modified organisms (GMOs) outside laboratory brings about safety concerns, it is necessary to explore the possibility of integrating a kill-switch. Toxin-Antitoxin (TA) systems were used to construct a mutually dependent host-plasmid platform and biocontainment systems in various academic and industrious situations. We selected nine TA systems from various bacteria strains and measured the toxicity of toxins (T) and the detoxifying activity of cognate antitoxins (A) to validate their potential to be used to build a kill-switch. These results prove the possibility of using engineered microorganisms to tackle antibiotic residues in environment efficiently and safely., Graphical abstract Image 1

3D geometric contents are becoming increasingly popular. In this paper, we study the problem of analyzing deforming 3D meshes using deep neural networks. Deforming 3D meshes are flexible to represent 3D animation sequences as well as collections of objects of the same category, allowing diverse shapes with large-scale non-linear deformations. We propose a novel framework which we call mesh variational autoencoders (mesh VAE), to explore the probabilistic latent space of 3D surfaces. The framework is easy to train, and requires very few training examples. We also propose an extended model which allows flexibly adjusting the significance of different latent variables by altering the prior distribution. Extensive experiments demonstrate that our general framework is able to learn a reasonable representation for a collection of deformable shapes, and produce competitive results for a variety of applications, including shape generation, shape interpolation, shape space embedding and shape exploration, outperforming state-of-the-art methods., Comment: CVPR 2018